freemo
155a5d7819
2.2 MiB
CleverAgents Documentation (Detailed Spec)
Overview
CleverAgents is your command center for AI agents—a unified platform for orchestrating any task you want agents to accomplish, from developing large software projects to writing comprehensive technical papers, administering databases, managing cloud infrastructure, or any complex multi-step workflow. The core value proposition is enabling long-running, complex, large-scale tasks to execute autonomously with minimal human intervention, making it ideal for building entire software systems, producing extensive documentation, or managing sophisticated operations largely hands-off.
In server mode, CleverAgents becomes a collaborative hub where teams can share resources—prompts, actors, actions, and projects—while executing plans in the cloud. This enables a consistent experience across all your devices: start a complex task on your laptop, check progress from your phone, and review results from any machine.
While CleverAgents leverages LangGraph and LangChain for the underlying LLM runtime primitives (tool calling, graphs, routing), its value lies in what it builds on top:
CleverAgents provides:
- A first-class plan lifecycle (Action templates driving Strategize/Execute/Apply phases) for breaking down and tracking complex work,
- A project + resource model for grounding tasks in real codebases, databases, documents, and infrastructure,
- A consistent actor abstraction for defining and composing intelligent agents,
- An independently registered resource abstraction for representing anything that can be read, written, or queried,
- An independently registered tool abstraction for reusable, callable operations with resource bindings,
- A validation abstraction that extends the tool concept with pass/fail semantics and resource-centric attachment with optional project/plan scoping,
- A consistent skill abstraction for organizing tools into composable capability collections,
- A sandbox + checkpoint safety model for safe, reversible execution,
- A CLI/TUI/Web UX for controlling and monitoring large multi-step autonomous work.
- A scalable Advanced Context Management System (ACMS) with a Universal Knowledge Ontology (UKO), a demand-driven Context Request Protocol (CRP), pluggable context strategies, a fusion coordinator, and hot/warm/cold tiers with per-actor views.
- Invariants as first-class constraints (global, project, action, and plan scoped) that flow into the decision tree, with precedence-based conflict resolution via the Invariant Reconciliation Actor.
- A future-facing correction model where the user can "edit the decision tree" and only recompute affected subtrees.
Glossary
- Plan: A ULID-identified, hierarchical unit of work instantiated from an Action template. Progresses through four phases — Action, Strategize, Execute, Apply — persisting a decision tree at each step. May spawn child plans. Scoped to one or more projects. Top-level plans (explicitly created by users) carry a namespaced name (
[[server:]namespace/]name); child plans (subplans spawned by a parent plan) are identified solely by their plan ID (ULID) and do not have a name. All plans are referenced by plan ID when precision is needed, especially in hierarchies containing subplans. - Action: A YAML-defined, reusable plan template specifying a description, definition of done, strategy/execution actors, typed arguments, and optional invariants. Project-agnostic until bound to one or more projects via
agents plan use, which instantiates a Plan. Namespaced as[[server:]namespace/]name. - Strategize: Second phase of the plan lifecycle (following the Action phase). The first phase where active processing occurs. Read-only: the strategy actor produces the initial decision tree — strategy choices, invariant enforcement records, resource selections, child plan blueprints — without modifying any resources.
- Execute: Third phase of the plan lifecycle. The execution actor carries out decisions from Strategize — invoking tools, spawning child plans, producing artifacts, creating checkpoints — with all mutations confined to a sandbox. May also create additional decisions constrained by Strategize-phase decisions, based on runtime discoveries. If Strategize constraints prove too restrictive, Execute reverts to Strategize for adjustment.
- Apply: Fourth and final phase of the plan lifecycle. Merges the sandbox changeset into real project resources. Terminal states within Apply:
applied(success — changes committed),constrained(cannot complete within current strategy constraints — may revert to Strategize),errored(failed),cancelled(user/system cancelled). If the plan enters theconstrainedstate, it may automatically or manually revert to Strategize for strategy adjustment. - Project: A named scope linking resources (from the Resource Registry), context policies, invariants, and validation attachments. Does not own resources; a single resource may be linked to multiple projects. Identified solely by its namespaced name (
[[server:]namespace/]name) — no ULID is generated. May be local or remote. - Resource: A ULID-identified entity registered in the Resource Registry representing anything readable, writable, or queryable (git repos, filesystems, databases, etc.). Classified as physical or virtual, typed by a Resource Type, and organized in a DAG via parent/child links. User-added resources also carry a namespaced name (
[[server:]namespace/]name); auto-discovered children have ULID only. - Resource Type: A schema-level definition constraining a category of resources. Specifies accepted CLI arguments, physical/virtual classification, permitted parent/child type relationships, auto-discovery rules, sandbox strategy, and handler implementation. Built-in types (e.g.,
git-checkout,fs-mount) are unnamespaced; custom types are YAML-defined and namespaced as[[server:]namespace/]name. - Physical Resource: A resource bound to a concrete, located artifact — a specific file at a specific path, a specific repo at a specific URL. Directly readable and writable by tools. Two physical resources with identical content remain distinct instances.
- Virtual Resource: A resource representing an abstract equivalence identity that links physical resources sharing the same content, hash, or name. Has no location; cannot be directly read or written. Equivalence relationships update when underlying physical resources diverge.
- Resource Binding: The association between a tool and the resources it operates on, declared via typed resource slots. Slots resolve through contextual binding (from the plan's project), static binding (hardcoded at registration), or parameter binding (passed at invocation).
- Resource Registry: The persistent catalog of all registered resources and their DAG relationships (parent/child links). One of the core registries, alongside the Tool Registry, Skill Registry, Actor Registry, and Provider Registry.
- Skill: A composable, namespaced collection of tools assembled by referencing named tools, defining inline anonymous tools, including other skills, or exposing MCP server / Agent Skills Standard tools. Actors reference skills by name to acquire capabilities. Namespaced as
[[server:]namespace/]name. - Tool: The atomic unit of execution: a namespaced, independently registered callable operation. Defined by JSON Schema inputs/outputs, capability metadata (
read_only,writes,checkpointable), and a four-stage lifecycle (discover/activate/execute/deactivate). Sources: MCP servers, Agent Skills Standard folders, built-ins, or custom Python. Namespaced as[[server:]namespace/]name. - Validation: A Tool subtype adding: a
mode(required|informational) controlling whether failure blocks execution; a structured JSON return with mandatorypassedboolean, optionaldata, and optionalmessage. Always read-only (writes = false,checkpointable = false). May wrap an existing Tool viawraps+transformto reuse its implementation. Shares the Tool naming namespace — no name collisions allowed. Usable anywhere a Tool is expected, but not vice versa. Managed viaagents validation add/attach/detach; listed, inspected, and removed through standardagents toolcommands. Always attached to a resource, optionally scoped to a project or plan. - Anonymous Tool: An inline tool definition embedded in a skill YAML or actor graph node. Same schema as a named tool but unregistered, unnamespaced, and scoped only to its defining context.
- Actor: A YAML-configured conversational unit — either a single LLM/agent or a composed LangGraph of actors and tool nodes. Specialized roles: strategy actor, execution actor, estimation actor, invariant reconciliation actor. Namespaced as
[[server:]namespace/]name. - Session: A persistent conversation thread tied to an orchestrator actor. Maintains message history across plans and serves as the user's natural-language interface.
- Server: An optional shared backend providing multi-user storage, namespace resolution, and remote plan execution.
- Namespace: The scoping segment in the name format
[[server:]namespace/]name. Defaults tolocal/when omitted.local/is reserved for local-only items (no server applies). Non-local/namespaces with server omitted assume the default configured server.<username>/and<orgname>/are server-stored. Built-in LLM actors use provider prefixes (e.g.,openai/,anthropic/). Built-in resource types are unnamespaced. - Decision: A persisted choice point in a plan's decision tree, created during Strategize or Execute. Records the question, chosen option, alternatives, confidence score, rationale, context snapshot, and downstream dependencies. Types:
prompt_definition,invariant_enforced,strategy_choice,subplan_spawn,subplan_parallel_spawn, among others. Execute-phase decisions are constrained by Strategize-phase decisions and arise from runtime discoveries. Supports targeted correction with selective subtree recomputation. - Invariant: A natural-language constraint on plan execution scoped to global, project, action, or plan level. Precedence: plan > action > project > global. Reconciled by the Invariant Reconciliation Actor at the start of Strategize; recorded as
invariant_enforceddecisions that propagate to child plans. - Automation Profile: A named set of confidence thresholds (each 0.0–1.0) gating which plan operations proceed automatically versus requiring human approval. Covers phase transitions, decision autonomy, self-repair, child plan spawning, and safety requirements. 0.0 = always automatic; 1.0 = always manual. Eight built-in profiles (
manualthroughfull-auto). Custom profiles namespaced as[[server:]namespace/]name. - ULID: Universally Unique Lexicographically Sortable Identifier. Assigned to plans, decisions, resources, correction attempts, and validation attachments. Projects, actions, skills, and tools use their namespaced name as sole identifier instead.
- ACMS (Advanced Context Management System): The pluggable, strategy-driven framework for assembling actor context. Comprises the UKO, CRP, pluggable context strategies, the Context Assembly Pipeline, and hot/warm/cold tiered storage with per-actor scoped views.
- UKO (Universal Knowledge Ontology): An RDF-based, inheritance-driven ontology representing resources at multiple abstraction levels with provenance and temporal versioning. Four layers: universal foundation, domain specializations (software, documents, data schemas, infrastructure), paradigm/format specializations (procedural programming, markdown), and technology-specific (Python, PostgreSQL). Semantically aware — implicit relationships are inferred from content analysis.
- CRP (Context Request Protocol): A structured vocabulary through which actors declare needed information, desired detail depth, and scope.
- Context Strategy: A pluggable retrieval component that searches for and assembles ContextFragments using a specific approach (keyword search, semantic embedding, graph navigation, temporal archaeology, etc.). Registered with the Context Assembly Pipeline; executed in parallel by the StrategyExecutor.
- Context Assembly Pipeline: The central ACMS orchestrator. Ten pluggable Protocol-defined components in three phases: Strategy Orchestration, Fragment Fusion, and Context Finalization. Each component ships with a default implementation and is overridable at global, project, or plan scope.
- Skeleton (context): A compressed representation of a plan's accumulated context produced by the SkeletonCompressor. Propagated from parent plans to child plans as inherited context. Size governed by the
skeleton_ratiobudget parameter.
CLI Commands
!!! adr "Architecture Decision" The CLI command structure, output rendering, and interaction patterns are defined in ADR-021: CLI and Output Rendering.
Command Synopsis
agents|cleveragents [--data-dir <DATA_PATH>] [--config-path <CONFIG_PATH>] [--format (rich|color|table|plain|json|yaml)] [--help|-h] [--version] [--install-completion [<INST_SHELL>]] [--show-completion [<SHOW_SHELL>]] [-v...] <COMMAND> [<ARGS>...]agents version agents info agents diagnostics agents init [--yes|-y]
agents session create [--actor <ACTOR>] agents session list agents session show <SESSION_ID> agents session delete [--yes|-y] <SESSION_ID> agents session export [(--output|-o) <FILE>] <SESSION_ID> agents session import (--input|-i) <FILE> agents session tell --session <SESSION_ID> [--actor <ACTOR>] [--stream] <PROMPT>
agents project create [(--description|-d) <DESC>] [--resource <RESOURCE>]... [--invariant <INVARIANT>]... [--invariant-actor <ACTOR>] <NAME> agents project link-resource [--read-only] <PROJECT> <RESOURCE> agents project unlink-resource [--yes|-y] <PROJECT> <RESOURCE_NAME> agents project list [(--namespace|-n) NS] [<REGEX>] agents project show <PROJECT> agents validation add (--config|-c) <FILE> [--update] agents validation attach [--project <PROJECT>|--plan <PLAN_ID>] <RESOURCE> <VALIDATION> [<ARGS>...] agents validation detach [--yes|-y] <ATTACHMENT_ID> agents project delete [--force|-f] [--yes|-y] <NAME> agents project context set[--view (strategize|execute|apply|default)] [--include-resource <INCLUDE_RESOURCE>]... [--exclude-resource <EXCLUDE_RESOURCE>]... [--include-path <INCLUDE_GLOB>]... [--exclude-path <EXCLUDE_GLOB>]... [--hot-max-tokens <N>] [--warm-max-decisions <N_WARM_MAX>] [--cold-max-decisions <N_COLD_MAX>] [--query-limit <N>] [--max-file-size <MAX_FILE_BYTES>] [--max-total-size <MAX_TOTAL_BYTES>] [--summarize|--no-summarize] [--summary-max-tokens <N>] [--strategy <STRATEGY>]... [--default-breadth <N>] [--default-depth <INT_OR_NAME>] [--depth-gradient <HOP:INT_OR_NAME>]... [--skeleton-ratio <FLOAT>] [--temporal-scope (current|recent|all)] [--auto-refresh|--no-auto-refresh] [--clear] <PROJECT> agents project context show [--view (strategize|execute|apply|default)] <PROJECT> agents project context inspect [--view (strategize|execute|apply|default)] [--strategy <STRATEGY>] [--focus <UKO_URI>]... [--breadth <N>] [--depth <INT_OR_NAME>] <PROJECT> agents project context simulate [--view (strategize|execute|apply|default)] [--budget <TOKENS>] [--focus <UKO_URI>]... [--strategy <STRATEGY>]... <PROJECT>
agents actor run [(--output|-o) <OUTPUT_FILE>] [-v...] [--unsafe|-u] [--context <CONTEXT_NAME>] [--context-dir <CONTEXT_PATH>] [--load-context <LOAD_CONTEXT_NAME>] [(--temperature|-t) <TEMP>] [--allow-rxpy-in-run-mode] [--skill <SKILL>]... <NAME> <PROMPT> agents actor add (--config|-c) <FILE> [--update] agents actor remove <NAME> agents actor list agents actor show <NAME> agents actor context remove [--yes|-y] (--all|-a|<NAME>) agents actor context list [<REGEX>] agents actor context show <NAME> agents actor context export (--output|-o) <FILE> <NAME> agents actor context import [--update] (--input|-i) <FILE> [<NAME>] agents actor context clear [--yes|-y] (--all|-a|<NAME>)
agents skill add (--config|-c) <FILE> [--update] agents skill remove [--yes|-y] <NAME> agents skill list [(--namespace|-n) <NS>] [--source <SOURCE>] agents skill show <NAME> agents skill tools <NAME>
agents tool add (--config|-c) <FILE> [--update] agents tool remove [--yes|-y] <NAME> agents tool list [(--namespace|-n) <NS>] [--source <SOURCE>] [--type (tool|validation)] [<REGEX>] agents tool show <NAME>
agents resource type add (--config|-c) <FILE> [--update] agents resource type remove [--yes|-y] <NAME> agents resource type list [<REGEX>] agents resource type show <NAME>
agents resource add [(--description|-d) <DESC>] [--update] <TYPE> <NAME> [type-specific-flags...] agents resource remove [--yes|-y] <NAME> agents resource list [--all] [(--type|-t) <TYPE>] agents resource show <RESOURCE> agents resource inspect [--tree] [--file <PATH>] <RESOURCE> agents resource tree [(--depth|-d) <N>] [(--type|-t) <TYPE>] <RESOURCE> agents resource link-child <PARENT> <CHILD> agents resource unlink-child [--yes|-y] <PARENT> <CHILD>
agents plan list [--phase <PHASE>] [--state <STATE>] [--project <PROJECT>] [--action <ACTION>] [<REGEX>] agents plan use [--automation-profile <PROFILE>] [--invariant <INVARIANT>]... [--strategy-actor <STRATEGY_ACTOR>] [--execution-actor <EXEC_ACTOR>] [--estimation-actor <EST_ACTOR>] [--invariant-actor <INV_ACTOR>] [--arg/-a name=value]... <ACTION> <PROJECT>... agents plan execute <PLAN_ID> agents plan apply [--yes|-y] <PLAN_ID> agents plan status <PLAN_ID> agents plan cancel [(--reason|-r) <REASON>] <PLAN_ID> agents plan tree [--show-superseded] <PLAN_ID> agents plan explain [--show-context] [--show-reasoning] <DECISION_ID> agents plan correct --mode (revert|append) (--guidance|-g) <GUIDANCE> [--dry-run] [--yes|-y] <DECISION_ID> agents plan diff (--correction <CORRECTION_ATTEMPT_ID>|<PLAN_ID>) agents plan artifacts <PLAN_ID> agents plan prompt <PLAN_ID> <GUIDANCE> agents plan rollback [--yes|-y] <PLAN_ID> <CHECKPOINT_ID>
agents action create (--config|-c) <CFG_FILE> agents action list [(--namespace|-n) <NS>] [(--state|-s) <STATE>] [<REGEX>] agents action show <ACTION_NAME> agents action archive <ACTION_NAME>
agents automation-profile add (--config|-c) <FILE> [--update] agents automation-profile remove [--yes|-y] <NAME> agents automation-profile list [<REGEX>] agents automation-profile show <NAME>
agents config set <key> <value> agents config get <key> agents config list [--filter-values <REGEX>] [<REGEX>]
agents invariant add [--global] [(--project|-p) PROJECT] [--plan PLAN_ID]... [--action ACTION]... <INVARIANT_TEXT> agents invariant list [--global] [(--project|-p) PROJECT] [--plan PLAN_ID] [--action ACTION] [--effective] [<REGEX>] agents invariant remove [--yes|-y] <INVARIANT_ID>
Command Reference
Global Options
Purpose Configure global state locations and shell integration for every command.
Arguments
--data-dir PATH: Overrides the global data directory (database, caches, sessions, logs). When omitted, the default data location is used.--config-path PATH: Overrides the global configuration file path. When omitted, the default config path is used.--format rich|color|table|plain|json|yaml: Set the output rendering format for all subcommands. When omitted, the value is read from the global config keycore.format. If not set in config either, defaults torich. See Output Rendering Framework for full details on each format. The available formats are:rich(default): Uses modern rich CLI elements with dynamic effects, animated spinners, progress bars, and generous color. Best for interactive terminal use.color: Plain scrolling text with ANSI color codes. Good for terminals that support color but not advanced rendering.table: ASCII box-drawing characters to create structured tables and panels with color. Similar to the examples shown throughout this document.plain: Plain text with no color codes or non-ASCII characters. Suitable for piping into files, logs, or non-terminal consumers.json: Structured JSON output for programmatic consumption. No color except within verbatim text values.yaml: Structured YAML output for programmatic consumption. No color except within verbatim text values.
--help,-h: Print help for the current command.--version: Print the version and exit.--install-completion [SHELL]: Install shell completion for the given shell.--show-completion [SHELL]: Show the completion script for the given shell.-v: Increase log verbosity for a single invocation (repeatable). By default (no-v), only normal command output is shown on stdout; logging output is suppressed except on fatal errors where the application exits entirely and displays the error. Each additional-vraises the verbosity one level:-v= ERROR (non-fatal, recoverable errors),-vv= WARN (warnings that may indicate a degraded system but are not necessarily errors),-vvv= INFO (routine informative messages useful to common users),-vvvv= DEBUG (coarse debugging output),-vvvvv= TRACE (the most granular logging level available). Increased verbosity writes to both the log file and stderr by default, keeping log output separate from normal program output on stdout. The destination for log output raised by-vis configurable: it can be sent to the terminal (stderr by default, or redirected to another stream such as stdin or a device), to the log file only, or to both. Seecore.log.terminal,core.log.terminal-stream, andcore.log.file-enabledconfiguration keys for details. This flag does not persist — it only affects the current invocation.
Examples
$ agents --data-dir /srv/cleveragents --config-path /srv/cleveragents/config.toml info╭─ System Snapshot ─────────────────╮ │ CleverAgents 1.0.0 │ │ Mode: local │ │ Automation: review │ │ Status: ready │ ╰───────────────────────────────────╯
╭─ Paths ───────────────────────────────╮ │ Data Dir: /srv/cleveragents │ │ Config: /srv/cleveragents/config.toml │ │ Logs: /srv/cleveragents/logs │ │ Cache: /srv/cleveragents/cache │ │ Database: /srv/cleveragents/agents.db │ ╰───────────────────────────────────────╯
╭─ Runtime ──────────╮ │ PID: 4127 │ │ Uptime: 00:14:32 │ │ Python: 3.13.1 │ │ Host: devbox.local │ │ Platform: linux │ ╰────────────────────╯
╭─ Projects & Sessions ─╮ │ Projects: 2 │ │ Sessions: 1 active │ │ Active Plans: 0 │ │ Actors: 3 │ ╰───────────────────────╯
✓ OK Environment loaded
agents version
agents versionPurpose Print the current CLI version.
Arguments
None.
Examples
$ agents version╭─ CLI Version ────╮ │ CleverAgents CLI │ │ Version: 1.0.0 │ │ Channel: stable │ │ Python: 3.13 │ ╰──────────────────╯
╭─ Build ────────────────╮ │ Build Date: 2026-02-08 │ │ Commit: a17c3f9 │ │ Schema: v3 │ │ Platform: linux-x86_64 │ ╰────────────────────────╯
╭─ Dependencies ─────────────────╮ │ LangGraph: 0.2.60 │ │ LangChain: 0.3.18 │ │ MCP SDK: 1.4.0 │ │ Pydantic: 2.10.4 │ ╰────────────────────────────────╯
✓ OK Version reported
agents info
agents infoPurpose Show configuration and runtime information useful for debugging or support.
Arguments
None.
Examples
$ agents info╭─ Environment ───────────────────────────────────────────────╮ │ Data Dir: /home/alex/.cleveragents │ │ Config: /home/alex/.cleveragents/config.toml │ │ Database: sqlite:///home/alex/.cleveragents/cleveragents.db │ │ Server Mode: disabled │ │ Platform: Linux 6.8.0 (x86_64) │ ╰─────────────────────────────────────────────────────────────╯
╭─ Runtime ─────────────────────────╮ │ Automation: review │ │ Providers: 3 configured │ │ Sessions: 2 active │ │ Active Plans: 1 │ ╰───────────────────────────────────╯
╭─ Storage ─────╮ │ Cache: 118 MB │ │ Logs: 42 MB │ │ Backups: 3 │ │ DB Size: 8 MB │ ╰───────────────╯
╭─ Indexing ─────────────╮ │ Text Index: ready │ │ Vector Index: ready │ │ Graph Store: disabled │ │ Indexed Files: 1,247 │ ╰────────────────────────╯
✓ OK Environment details ready
agents diagnostics
agents diagnosticsPurpose Run health checks for configuration, providers, and filesystem permissions.
Arguments
None.
Examples
$ agents diagnostics╭─ Checks ────────────────────────────────╮ │ Check Status Details │ │ ─────────────── ────── ────────────── │ │ Config file OK readable │ │ Database OK writable │ │ OPENAI_API_KEY WARN missing │ │ Anthropic key OK configured │ │ Disk space OK 2.1 GB free │ │ Text index OK tantivy 0.22 │ │ Vector index OK faiss (CPU) │ │ Graph store WARN not configured │ │ File permissions OK data dir r/w │ │ Git OK git 2.43.0 │ ╰─────────────────────────────────────────╯
╭─ Summary ─────────╮ │ Checks: 10 total │ │ Warnings: 2 │ │ Errors: 0 │ │ Duration: 0.6s │ ╰───────────────────╯
╭─ Recommendations ─────────────────────────────────────────────╮ │ - Set OPENAI_API_KEY to enable OpenAI models │ │ - Configure a graph store backend for structural code queries │ │ - Verify provider credentials via config │ ╰───────────────────────────────────────────────────────────────╯
⚠ WARN 2 warnings require attention
When critical checks fail, diagnostics reports errors:
$ agents diagnostics╭─ Checks ────────────────────────────────────────────╮ │ Check Status Details │ │ ─────────────── ─────── ──────────────────── │ │ Config file OK readable │ │ Database ERROR locked by another process │ │ OPENAI_API_KEY OK configured │ │ Anthropic key ERROR invalid key format │ │ Disk space WARN 312 MB free (low) │ │ Text index OK tantivy 0.22 │ │ Vector index ERROR FAISS library not found │ │ Graph store OK neo4j 5.15 │ │ File permissions OK data dir r/w │ │ Git OK git 2.43.0 │ ╰─────────────────────────────────────────────────────╯
╭─ Summary ─────────╮ │ Checks: 10 total │ │ Warnings: 1 │ │ Errors: 3 │ │ Duration: 1.2s │ ╰───────────────────╯
╭─ Errors (must fix) ─────────────────────────────────────────────────╮ │ 1. Database is locked by PID 12847 — stop the other process or │ │ delete the lock file at ~/.cleveragents/agents.db-lock │ │ 2. Anthropic key starts with "pk-" — expected "sk-ant-" prefix │ │ Run: agents config set provider.anthropic.api-key │ │ 3. FAISS library not installed — vector search will not work │ │ Run: pip install faiss-cpu │ ╰─────────────────────────────────────────────────────────────────────╯
✗ ERROR 3 errors must be resolved before CleverAgents can operate
agents init
agents init [--yes|-y]Purpose Initialize or reset the global CleverAgents environment. This wipes any existing data and re-creates the global config and database.
Arguments
--yes: Skip the confirmation prompt and proceed with the wipe.
Examples
$ agents initWarning: This will remove all data in /home/alex/.cleveragents Continue? [y/N]: y
╭─ Environment Reset ────────────────────────────────╮ │ Config: /home/alex/.cleveragents/config.toml │ │ Database: /home/alex/.cleveragents/cleveragents.db │ │ Backup: /home/alex/.cleveragents.backup-2026-02-08 │ │ Status: ready │ ╰────────────────────────────────────────────────────╯
╭─ Defaults ──────────────────────────────────╮ │ Automation Profile: supervised │ │ Built-in Profiles: 8 loaded │ ╰─────────────────────────────────────────────╯
╭─ Created ─────────────────╮ │ Config: config.toml │ │ Database: cleveragents.db │ │ Logs: logs/ │ │ Cache: cache/ │ │ Backups: backups/ │ ╰───────────────────────────╯
╭─ Schema ───────────────────────╮ │ Version: v3 │ │ Tables: 12 created │ │ Migrations: up to date │ ╰────────────────────────────────╯
✓ OK Environment initialized
Non-interactive initialization using --yes (useful in scripts and CI):
$ agents init --yes╭─ Initialized ──────────────────────────────────────────╮ │ Data Dir: /home/alex/.cleveragents (created) │ │ Config: /home/alex/.cleveragents/config.toml │ │ Database: initialized (schema v3) │ │ Directories: logs, cache, sessions, contexts │ ╰────────────────────────────────────────────────────────╯
✓ OK Initialized (non-interactive)
agents session
Purpose Manage interactive sessions that hold a conversation history and orchestrator state.
agents session create
agents session create [--actor <ACTOR>]Purpose Create a new session for interactive work.
Arguments
--actor ACTOR: Orchestrator actor to use forsession tell.
Examples
$ agents session create --actor local/orchestrator╭─ Session ───────────────────────╮ │ ID: 01HXM2A6K1P2E9Q9D4GQ7J4S7Z │ │ Actor: local/orchestrator │ │ Created: 2026-02-08 12:44 │ │ Namespace: local │ ╰─────────────────────────────────╯
╭─ Settings ─────────────╮ │ Automation: review │ │ Streaming: off │ │ Context: default │ │ Memory: enabled │ │ Max History: 50 turns │ ╰────────────────────────╯
╭─ Actor Details ───────────────────╮ │ Provider: anthropic │ │ Model: claude-3.5 │ │ Temperature: 0.7 │ │ Context Window: 200K tokens │ ╰───────────────────────────────────╯
✓ OK Session created
agents session list
agents session listPurpose List sessions available on the local machine.
Arguments
None.
Examples
$ agents --format table session list╭─ Sessions ───────────────────────────────────────────────────────────────────╮ │ ID Name Actor Messages Updated │ │ ──────── ─────────────── ────────────────── ──────── ──────────────── │ │ 01HXM2A6 weekly-planning local/orchestrator 6 2026-02-08 12:44 │ │ 01HXM1F2 refactor-sprint local/orchestrator 14 2026-02-07 18:11 │ ╰──────────────────────────────────────────────────────────────────────────────╯
╭─ Summary ────────────────────╮ │ Total: 2 │ │ Most Recent: weekly-planning │ │ Oldest: refactor-sprint │ │ Total Messages: 20 │ │ Storage: 42 KB │ ╰──────────────────────────────╯
✓ OK 2 sessions listed
agents session show
agents session show <SESSION_ID>Purpose Show details and recent messages for a session.
Arguments
<SESSION_ID>: The session identifier.
Examples
$ agents session show 01HXM2A6K1P2E9Q9D4GQ7J4S7Z╭─ Session Summary ───────────────╮ │ ID: 01HXM2A6K1P2E9Q9D4GQ7J4S7Z │ │ Actor: local/orchestrator │ │ Messages: 6 │ │ Created: 2026-02-08 12:30 │ │ Updated: 2026-02-08 12:44 │ │ Automation: review │ ╰─────────────────────────────────╯
╭─ Recent Messages ──────────────────────────────────╮ │ user Create an action to refresh dependency locks │ │ assistant Plan created, running commands... │ │ assistant Completed 2 commands │ ╰────────────────────────────────────────────────────╯
╭─ Linked Plans ────────────────────────────────╮ │ Plan ID Phase State │ │ ────────────────────────── ────── ──────── │ │ 01HXM8C2ZK4Q7C2B3F2R4VYV6J execute complete │ ╰───────────────────────────────────────────────╯
╭─ Token Usage ──────────────╮ │ Input Tokens: 3,420 │ │ Output Tokens: 1,185 │ │ Estimated Cost: $0.0184 │ ╰────────────────────────────╯
✓ OK Session details loaded
agents session delete
agents session delete [--yes|-y] <SESSION_ID>Purpose Delete a session and its stored conversation history.
Arguments
<SESSION_ID>: The session identifier.--yes, -y: Skip the confirmation prompt.
Examples
$ agents session delete 01HXM2A6K1P2E9Q9D4GQ7J4S7ZDelete session 01HXM2A6K1P2E9Q9D4GQ7J4S7Z? [y/N]: y
╭─ Deletion Summary ──────────────────╮ │ Session: 01HXM2A6K1P2E9Q9D4GQ7J4S7Z │ │ ID: 01HXM2A6K1P2E9Q9D4GQ7J4S7Z │ │ Messages: 6 removed │ │ Storage: 18 KB freed │ │ Plans Orphaned: 0 │ ╰─────────────────────────────────────╯
╭─ Cleanup ───────────╮ │ Backups: none │ │ Logs: preserved │ │ Context: cleared │ │ Checkpoints: none │ ╰─────────────────────╯
✓ OK Session deleted
agents session export
agents session export [(--output|-o) <FILE>] <SESSION_ID>Purpose Export a session as a portable JSON file.
Arguments
<SESSION_ID>: The session identifier.--output/-o FILE: Output file path (optional).
Examples
$ agents session export --output /tmp/weekly-planning.json 01HXM2A6K1P2E9Q9D4GQ7J4S7Z╭─ Session Export ────────────────────╮ │ Session: 01HXM2A6K1P2E9Q9D4GQ7J4S7Z │ │ Output: /tmp/weekly-planning.json │ │ Messages: 6 │ │ Size: 24 KB │ │ Format: JSON │ ╰─────────────────────────────────────╯
╭─ Contents ─────────────────╮ │ Messages: 6 │ │ Plan References: 1 │ │ Metadata Keys: 2 │ │ Actor Config: included │ │ Schema Version: v3 │ ╰────────────────────────────╯
╭─ Integrity ──────────────────╮ │ Checksum: sha256:7a9b...42c1 │ │ Encrypted: no │ ╰──────────────────────────────╯
✓ OK Export completed
agents session import
agents session import (--input|-i) <FILE>Purpose Import a session JSON file.
Arguments
--input/-i FILE: Input JSON file.
Examples
$ agents session import --input /tmp/weekly-planning.json╭─ Session Import ────────────────────────╮ │ Input: /tmp/weekly-planning.json │ │ Session ID: 01HXM3D3B2W4CQYQ3P4ZB8A5T1 │ │ Messages: 6 │ │ Schema: v3 │ ╰─────────────────────────────────────────╯
╭─ Validation ────────────╮ │ Checksum: verified │ │ Schema: compatible │ │ Actor Ref: resolved │ ╰─────────────────────────╯
╭─ Merge ──────────────╮ │ Existing: none │ │ Strategy: create new │ ╰──────────────────────╯
✓ OK Import completed
agents session tell
agents session tell --session <SESSION_ID> [--actor <ACTOR>] [--stream] <PROMPT>Purpose Send a natural-language request to the orchestrator. The orchestrator can create actions, plans, or project changes by issuing the necessary CleverAgents commands under the hood.
Arguments
<PROMPT>: Instruction text (positional argument).--session SESSION_ID: Session to use (required).--actor ACTOR: Override the session actor for this request.--stream: Stream progress as the orchestrator works.
Examples
$ agents session tell "Create an action to refresh dependency locks and add it to the platform project" \ --session 01HXM2A6K1P2E9Q9D4GQ7J4S7Z╭─ Plan Request ──────────────────────────────────────────╮ │ Actor: local/orchestrator │ │ Session: 01HXM2A6K1P2E9Q9D4GQ7J4S7Z │ │ Automation: review │ │ Prompt: Create an action to refresh dependency locks... │ ╰─────────────────────────────────────────────────────────╯
╭─ Commands Executed ─────────────────────────────────────────────────────────────────────────────────────────╮ │ - agents action create --config ./actions/refresh-locks.yaml │ │ - agents resource add git-checkout local/platform-repo --path /repos/platform │ │ - agents project link-resource local/platform local/platform-repo │ ╰─────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Result ──────────────────────────╮ │ Action: local/refresh-locks │ │ Project: local/platform │ │ Resource: local/platform-repo │ ╰───────────────────────────────────╯
╭─ Usage ─────────────────────╮ │ Input Tokens: 1,842 │ │ Output Tokens: 624 │ │ Cost: $0.0094 │ │ Duration: 3.2s │ │ Tool Calls: 3 │ ╰─────────────────────────────╯
✓ OK Orchestrator completed 3 commands
Using --stream to see the response as it is generated (token by token):
$ agents session tell --session 01HXM2A6K1 --stream "What files were changed in the last plan?"╭─ Session ──────────────────────────╮ │ ID: 01HXM2A6K1P2E9Q9D4GQ7J4S7Z │ │ Actor: local/orchestrator │ │ Mode: streaming │ ╰────────────────────────────────────╯
▍ The last plan (01HXM8C2ZK) modified 6 files in the auth module:
src/auth/session.py— refactored session validationsrc/auth/tokens.py— updated token expiry to 7200ssrc/auth/__init__.py— updated exportstests/test_auth.py— added 12 new test casestests/test_session.py— updated session fixturesdocs/auth.md— updated API documentationAll changes passed validation (24/24 tests, lint clean).
╭─ Usage ──────────────────╮ │ Tokens: 1,240 (stream) │ │ Duration: 3.1s │ │ Tool Calls: 2 │ ╰──────────────────────────╯
✓ OK Stream complete
agents project
Purpose Manage projects and their resources.
agents project create
agents project create [(--description|-d) <DESC>] [--resource <RESOURCE>]...
[--invariant <INVARIANT>]... [--invariant-actor <ACTOR>] <NAME>Purpose Create a new project record.
Arguments
<NAME>: Namespaced project name (positional argument).--description/-d TEXT: Optional description.--resource RESOURCE: Resource name or ULID to link to the project at creation time (repeatable).--invariant TEXT: Invariant to attach to this project (repeatable). These invariants apply to all plans targeting this project.--invariant-actor ACTOR: Invariant Reconciliation Actor for this project. Used to reconcile project-level invariants against global invariants for all plans targeting this project (unless the plan overrides it).
Examples
$ agents project create --description "Backend API" local/api-service╭─ Project ──────────────────────╮ │ Name: local/api-service │ │ Description: Backend API │ │ Type: local │ │ Created: 2026-02-08 12:46 │ ╰────────────────────────────────╯
╭─ Paths ────────────────────────────────────╮ │ Root: /repos/api-service │ │ Data Dir: /repos/api-service/.cleveragents │ ╰────────────────────────────────────────────╯
╭─ Defaults ──────────────────────────────╮ │ Sandbox: git_worktree │ │ Validations: 0 │ │ Context Filters: none │ │ Automation Profile: (inherits global) │ ╰─────────────────────────────────────────╯
╭─ Resources ──────╮ │ Total: 0 │ │ Indexed: 0 │ │ Sandboxable: 0 │ ╰──────────────────╯
✓ OK Project created
Creating a project with resources and invariants in one command:
$ agents project create -d "Frontend web application" \ --resource local/web-repo --resource local/web-db \ --invariant "All components must have unit tests" \ --invariant "CSS must pass stylelint checks" \ --invariant-actor local/invariant-resolver \ local/web-app╭─ Project Created ─────────────────────╮ │ Name: local/web-app │ │ Description: Frontend web app │ │ Remote: no │ ╰───────────────────────────────────────╯
╭─ Linked Resources ───────────────────────────────────────╮ │ Name Type Read-Only │ │ ───────────────── ────────────── ───────── │ │ local/web-repo git-checkout no │ │ local/web-db local/database no │ ╰──────────────────────────────────────────────────────────╯
╭─ Invariants ────────────────────────────────────╮ │ 1. All components must have unit tests │ │ 2. CSS must pass stylelint checks │ │ Reconciliation Actor: local/invariant-resolver │ ╰─────────────────────────────────────────────────╯
✓ OK Project created
agents project link-resource
agents project link-resource [--read-only] <PROJECT> <RESOURCE>Purpose
Link a registered resource to a project. The resource must already be registered via agents resource add. A resource can be linked to multiple projects. Optionally, the resource can be marked as read-only within this project context, and given an alias for convenience.
Arguments
<PROJECT>: Project name (positional argument).<RESOURCE>: Resource name or ULID (positional argument).--read-only: Mark the resource as read-only within this project (even if the resource itself is writable).
Examples
$ agents resource add git-checkout local/api-repo --path /repos/api --branch main $ agents project link-resource local/api-service local/api-repo╭─ Resource Linked ───────────────────────╮ │ Project: local/api-service │ │ Resource: local/api-repo │ │ Type: git-checkout │ │ Read-Only: no │ ╰─────────────────────────────────────────╯
╭─ Access ─────────────────╮ │ Read: allowed │ │ Write: allowed │ │ Apply: requires approval │ ╰──────────────────────────╯
╭─ Indexing ─────────────────────╮ │ Status: indexing... │ │ Files Found: 347 │ │ Language: Python (primary) │ │ Estimated Time: ~20 seconds │ ╰────────────────────────────────╯
✓ OK Resource linked to project
agents project unlink-resource
agents project unlink-resource [--yes|-y] <PROJECT> <RESOURCE_NAME>Purpose Unlink a resource from a project. The resource itself remains registered in the Resource Registry and can still be linked to other projects. Active plans using this resource in this project context will be warned.
Arguments
<PROJECT>: Project name (positional argument).<RESOURCE_NAME>: Resource name (positional argument). Only user-added (named) resources can be unlinked.--yes, -y: Skip confirmation prompt.
Examples
$ agents project unlink-resource local/api-service local/api-repoUnlink local/api-repo from local/api-service? [y/N]: y
╭─ Resource Unlinked ─────────────────────────────────╮ │ Project: local/api-service │ │ Resource: local/api-repo │ │ Type: git-checkout │ ╰─────────────────────────────────────────────────────╯
╭─ Index Cleanup ──────────╮ │ Text Index: 347 removed │ │ Vectors: 892 removed │ │ Graph Triples: cleared │ │ Duration: 0.3s │ ╰──────────────────────────╯
╭─ Project Summary ──────────────╮ │ Linked Resources: 1 remaining │ │ Last Updated: 2026-02-09 10:30 │ │ Active Plans: 0 │ ╰────────────────────────────────╯
✓ OK Resource unlinked from project
agents project list
agents project list [(--namespace|-n) NS] [<REGEX>]Purpose List projects with optional filters.
Arguments
--namespace/-n NS: Filter by namespace.
Examples
$ agents --format table project list╭─ Projects ────────────────────────────────────────────────────╮ │ Name Resources Remote Active Plans │ │ ───────────────── ───────── ────── ──────────── │ │ local/api-service 2 No 1 │ │ local/docs 1 No 0 │ ╰───────────────────────────────────────────────────────────────╯
╭─ Summary ──────────────────╮ │ Total: 2 │ │ With Resources: 2 │ │ Remote: 0 │ │ Total Resources: 3 │ │ Indexed Files: 1,247 │ │ Active Plans: 1 │ ╰────────────────────────────╯
✓ OK 2 projects listed
agents project show
agents project show <PROJECT>Purpose Show full project details.
Arguments
<NAME>: Project name.
Examples
$ agents project show local/api-service╭─ Project Details ──────────────╮ │ Name: local/api-service │ │ Description: Backend API │ │ Resources: 2 │ │ Remote: no │ │ Created: 2026-02-08 12:46 │ ╰────────────────────────────────╯
╭─ Linked Resources ──────────────────────────────────────────────────────╮ │ Resource Type Sandbox Read-Only │ │ ──────────────── ────────────── ──────────────────── ───────── │ │ local/api-repo git-checkout git_worktree no │ │ local/staging-db local/database transaction_rollback yes │ ╰─────────────────────────────────────────────────────────────────────────╯
╭─ Validations (3) ──────────────────────────────────────────────────────────────────╮ │ local/run-tests Run unit tests with coverage required │ │ resource: local/api-repo scope: project (attachment: 01HXM5A1B2C3D4E5F6G7…)│ │ local/lint-check Lint check required │ │ resource: local/api-repo scope: direct (always active) (attachment: 01HXM5…)│ │ local/check-bundle-size Check bundle size (advisory) info │ │ resource: local/api-repo scope: project (attachment: 01HXM5D4E5F6G7H8J9…)│ ╰───────────────────────────────────────────────────────────────────────────────────╯
╭─ Context ───────────────────╮ │ Include: repo │ │ Exclude: /node_modules/ │ │ Max File Size: 1 MB │ ╰─────────────────────────────╯
╭─ Indexing Status ──────────╮ │ Text Index: ready │ │ Vector Index: ready │ │ Graph Store: disabled │ │ Indexed Files: 347 │ │ Last Indexed: 12:48 │ ╰────────────────────────────╯
╭─ Active Plans ──────────────────────────╮ │ Plan ID Action Phase │ │ ──────── ─────────────────── ─────── │ │ 01HXM7A9 local/code-coverage execute │ ╰─────────────────────────────────────────╯
✓ OK Project loaded
agents project validations
Note: Validations are managed as a top-level entity via agents validation commands (see the agents validation section below). A validation is always attached to a resource, but can optionally be scoped to a project via agents validation attach --project <PROJECT> <RESOURCE> <VALIDATION>. When scoped to a project, the validation only runs when that resource is accessed through the specified project. The agents project show command displays all validations relevant to the project — both directly-attached validations on linked resources and project-scoped validations.
See: agents validation for full details on creating, managing, and attaching validations.
agents project delete
agents project delete [--force|-f] [--yes|-y] <NAME>Purpose Delete a project and unlink all associated resources. The resources themselves remain in the Resource Registry.
Arguments
<NAME>: Project name.--force/-f: Delete even if active plans exist.--yes: Skip confirmation prompt.
Examples
$ agents project delete local/docsDelete project local/docs? This cannot be undone. [y/N]: y
╭─ Deletion Summary ──────────────────╮ │ Project: local/docs │ │ Resources: 1 unlinked │ │ Data Dir: /repos/docs/.cleveragents │ ╰─────────────────────────────────────╯
╭─ Index Cleanup ────────╮ │ Text Index: cleared │ │ Vectors: 240 removed │ │ Graph Triples: none │ │ Storage Freed: 12 MB │ ╰────────────────────────╯
╭─ Backups ────────────────────────────────────╮ │ Snapshot: /backups/local-docs-2026-02-08.tgz │ │ Retention: 7 days │ ╰──────────────────────────────────────────────╯
✓ OK Project deleted
Attempting to delete a project that has active plans (without --force):
$ agents project delete local/api-serviceDelete project local/api-service? [y/N]: y
╭─ Delete Blocked ──────────────────────────────────────────╮ │ Cannot delete: project has active plans │ │ Active Plans: 2 │ │ 01HXM7A9 — execute/processing (local/code-coverage) │ │ 01HXM4J1 — strategize/processing (local/refactor-api) │ ╰───────────────────────────────────────────────────────────╯
╭─ Resolution ──────────────────────────────────────────────────────╮ │ - Cancel or complete active plans first, or │ │ - Use --force to cancel all active plans and delete the project │ ╰───────────────────────────────────────────────────────────────────╯
✗ ERROR Delete blocked — 2 active plans
Force-deleting a project with active plans:
$ agents project delete --force --yes local/api-service╭─ Force Delete ─────────────────────────────────────────╮ │ Cancelling: 2 active plans │ │ 01HXM7A9 — cancelled (was execute/processing) │ │ 01HXM4J1 — cancelled (was strategize/processing) │ ╰────────────────────────────────────────────────────────╯
╭─ Deleted ────────────────────╮ │ Project: local/api-service │ │ Resources Unlinked: 2 │ │ Validations Detached: 3 │ │ Plans Cancelled: 2 │ │ Context Policies: cleared │ ╰──────────────────────────────╯
✓ OK Project force-deleted
agents project context
Purpose Manage ACMS context policies for the hot/warm/cold tiers, per-view context selection, strategy configuration, UKO depth/breadth parameters, and context budget tuning.
agents project context set
agents project context set[--view (strategize|execute|apply|default)]
[--include-resource <INCLUDE_RESOURCE>]...
[--exclude-resource <EXCLUDE_RESOURCE>]...
[--include-path <INCLUDE_GLOB>]...
[--exclude-path <EXCLUDE_GLOB>]...
[--hot-max-tokens <N>]
[--warm-max-decisions <N_WARM_MAX>]
[--cold-max-decisions <N_COLD_MAX>]
[--query-limit <N>]
[--max-file-size <MAX_FILE_BYTES>]
[--max-total-size <MAX_TOTAL_BYTES>]
[--summarize|--no-summarize]
[--summary-max-tokens <N>]
[--strategy <STRATEGY>]...
[--default-breadth <N>]
[--default-depth <INT_OR_NAME>]
[--depth-gradient <HOP:INT_OR_NAME>]...
[--skeleton-ratio <FLOAT>]
[--temporal-scope (current|recent|all)]
[--auto-refresh|--no-auto-refresh]
[--clear] <PROJECT>Purpose Set the context policy for a project and (optionally) a specific view.
Arguments
<PROJECT>: Project name (positional argument, at end of command).--view strategize|execute|apply|default: Which view this policy applies to.--include-resource NAME: Resource allowlist (repeatable).--exclude-resource NAME: Resource denylist (repeatable).--include-path GLOB: Path allowlist (repeatable).--exclude-path GLOB: Path denylist (repeatable).--hot-max-tokens N: Maximum token budget for hot context. This is a soft cap and may be null. The actor/LLM hard limit can be lower; the effective hot context is the lesser of the two.--warm-max-decisions N: Maximum number of decisions kept in warm context.--cold-max-decisions N: Maximum number of decisions kept in cold context.--query-limit N: Max number of retrieval results per query.--max-file-size BYTES: Max file size included in context.--max-total-size BYTES: Max total size across included files.--summarize/--no-summarize: Enable or disable summarization for large context segments.--summary-max-tokens N: Token limit for generated summaries.--strategy STRATEGY: ACMS context strategy to enable for this view (repeatable). Overrides the globalcontext.strategies.enabledlist. Built-in:simple-keyword,semantic-embedding,breadth-depth-navigator,arce,temporal-archaeology,plan-decision-context.--default-breadth N: Default number of hops from focus nodes in the UKO graph.0= focus nodes only. Default:2.--default-depth INT_OR_NAME: Default detail depth for context fragments. Accepts a raw integer (e.g.,4) or a named level from the active domain's DetailLevelMap (e.g.,SIGNATURES,MEMBER_SUMMARY). Named levels are resolved to integers via the DetailLevelMap inheritance chain. Default:3(typicallyMEMBER_SUMMARY/FULL_TOC/TYPED_COLUMNSdepending on domain).--depth-gradient HOP:INT_OR_NAME: Per-hop detail depth override (repeatable). Format:0:9,1:4,2:0or0:FULL_SOURCE,1:SIGNATURES,2:MODULE_LISTING. Hops not specified use--default-depth.--skeleton-ratio FLOAT: Fraction of context budget reserved for inherited plan skeleton. Range: 0.0–1.0. Default:0.15.--temporal-scope current|recent|all: Temporal scope for UKO node resolution. Default:current.--auto-refresh/--no-auto-refresh: Enable or disable automatic context re-assembly on budget changes.--clear: Clear the policy for the selected view.
Examples
$ agents project context set --view strategize \ --include-resource repo --exclude-path "**/node_modules/**" \ --hot-max-tokens 12000 --warm-max-decisions 50 --cold-max-decisions 200 \ --summarize --summary-max-tokens 800 local/api-service╭─ Context Policy ────────────╮ │ Project: local/api-service │ │ View: strategize │ │ Include: repo │ │ Exclude: /node_modules/ │ ╰─────────────────────────────╯
╭─ Limits ─────────────────────╮ │ Hot Tokens: 12000 (soft cap) │ │ Warm Decisions: 50 │ │ Cold Decisions: 200 │ │ Query Limit: 20 │ │ Max File Size: 1 MB │ │ Max Total Size: 50 MB │ ╰──────────────────────────────╯
╭─ Summarization ─╮ │ Enabled: yes │ │ Max Tokens: 800 │ ╰─────────────────╯
╭─ Other Views ────────╮ │ execute: (default) │ │ apply: (default) │ │ default: (unset) │ ╰──────────────────────╯
✓ OK Context policy updated
agents project context show
agents project context show [--view (strategize|execute|apply|default)]
<PROJECT>Purpose Show the active context policy for a project.
Arguments
<PROJECT>: Project name (positional argument, at end of command).--view strategize|execute|apply|default: View to display.
Examples
$ agents project context show --view strategize local/api-service╭─ Context Policy ────────────╮ │ Project: local/api-service │ │ View: strategize │ │ Include: repo │ │ Exclude: /node_modules/ │ ╰─────────────────────────────╯
╭─ Limits ─────────────────────╮ │ Hot Tokens: 12000 (soft cap) │ │ Warm Decisions: 50 │ │ Cold Decisions: 200 │ │ Query Limit: 20 │ │ Max File Size: 1 MB │ │ Max Total Size: 50 MB │ ╰──────────────────────────────╯
╭─ Summarization ─╮ │ Enabled: yes │ │ Max Tokens: 800 │ ╰─────────────────╯
╭─ Current Usage ──────────────╮ │ Hot Context: 8,420 / 12,000 │ │ Warm Entries: 12 / 50 │ │ Cold Entries: 47 / 200 │ │ Indexed Resources: 1 │ ╰──────────────────────────────╯
✓ OK Context policy loaded
agents project context inspect
agents project context inspect [--view (strategize|execute|apply|default)]
[--strategy <STRATEGY>]
[--focus <UKO_URI>]...
[--breadth <N>] [--depth <INT_OR_NAME>]
<PROJECT>Purpose Inspect the current state of the ACMS context assembly for a project. Shows the UKO graph structure, active strategies, indexed resources, context budget usage, and the most recent fusion result. Useful for debugging context quality issues.
Arguments
<PROJECT>: Project name (positional argument, at end of command).--view strategize|execute|apply|default: View to inspect. Default:default.--strategy STRATEGY: Filter to show results from a specific strategy only.--focus UKO_URI: Show the UKO subgraph rooted at these focus nodes (repeatable). When omitted, shows a summary of the full indexed graph.--breadth N: Override the breadth (hop count) for the focus node expansion. Default: uses the view'sdefault_breadth.--depth INT_OR_NAME: Override the detail depth for the display. Accepts a raw integer or a named level from the active DetailLevelMap. Default: uses the view'sdefault_depth.
Examples
$ agents project context inspect --view execute --focus uko-py:module/src.auth --breadth 2 local/api-service╭─ ACMS Context Inspection ────────────────────────────────────────╮ │ Project: local/api-service │ │ View: execute │ │ Focus: uko-py:module/src.auth │ │ Breadth: 2 hops │ Depth: 3 (MEMBER_SUMMARY) │ ╰──────────────────────────────────────────────────────────────────╯
╭─ UKO Graph (2-hop neighborhood) ─────────────────────────────────╮ │ uko-py:module/src.auth (depth 9/FULL_SOURCE, 1,240 tokens) │ │ ├─ contains → uko-py:class/AuthHandler (depth 3, 320 tokens) │ │ ├─ contains → uko-py:class/TokenValidator (depth 3, 280 tok) │ │ ├─ imports → uko-py:module/src.db (depth 0, 90 tokens) │ │ └─ imports → uko-py:module/src.config (depth 0, 60 tok) │ ╰──────────────────────────────────────────────────────────────────╯
╭─ Active Strategies ──────────────────────────────────────────────╮ │ breadth-depth-navigator quality=0.85 budget=45% 4 fragments │ │ semantic-embedding quality=0.60 budget=35% 6 fragments │ │ simple-keyword quality=0.30 budget=20% 3 fragments │ ╰──────────────────────────────────────────────────────────────────╯
╭─ Budget ─────────────────────────────╮ │ Model Window: 128,000 tokens │ │ Response Reserve: 4,096 tokens │ │ Tool Definitions: 2,400 tokens │ │ System Prompt: 1,200 tokens │ │ Skeleton Reserve: 18,046 (15%) │ │ Available Budget: 102,258 tokens │ │ Used: 1,990 / 102,258 (1.9%) │ ╰──────────────────────────────────────╯
✓ OK Context inspection complete
agents project context simulate
agents project context simulate [--view (strategize|execute|apply|default)]
[--budget <TOKENS>]
[--focus <UKO_URI>]...
[--strategy <STRATEGY>]...
<PROJECT>Purpose Simulate a context assembly run without invoking an actor. Produces a dry-run report showing what the Context Assembly Pipeline would include in the context payload, which strategies contributed, how the budget was allocated, and what was excluded. Useful for tuning context view configurations before running plans.
Arguments
<PROJECT>: Project name (positional argument, at end of command).--view strategize|execute|apply|default: View to simulate. Default:default.--budget TOKENS: Override the context budget in tokens. When omitted, uses the dynamic budget calculation.--focus UKO_URI: Focus nodes for the simulation (repeatable). When omitted, uses a default focus derived from recent plan activity.--strategy STRATEGY: Override which strategies to run (repeatable). When omitted, uses the view's configured strategies.
Examples
$ agents project context simulate --view strategize --budget 16000 \ --focus uko-py:module/src.auth --focus uko-py:module/src.db \ local/api-service╭─ ACMS Simulation ────────────────────────────────────────────────╮ │ Project: local/api-service │ │ View: strategize │ │ Budget: 16,000 tokens (override) │ │ Focus: uko-py:module/src.auth, uko-py:module/src.db │ ╰──────────────────────────────────────────────────────────────────╯
╭─ Strategy Results ───────────────────────────────────────────────╮ │ breadth-depth-navigator │ │ Confidence: 0.85 │ Budget: 7,200 tokens │ 12 fragments │ │ Top: src/auth/__init__.py (depth 9, 1,240t, score=0.95) │ │ src/db/models.py (depth 3, 480t, score=0.88) │ │ │ │ semantic-embedding │ │ Confidence: 0.60 │ Budget: 5,600 tokens │ 8 fragments │ │ Top: src/auth/jwt.py (depth 3, 380t, score=0.82) │ │ src/middleware/auth_check.py (depth 3, 290t, score=0.79) │ │ │ │ simple-keyword │ │ Confidence: 0.30 │ Budget: 3,200 tokens │ 5 fragments │ │ Top: src/auth/utils.py (depth 6, 210t, score=0.71) │ ╰──────────────────────────────────────────────────────────────────╯
╭─ Fusion Result ──────────────────────────────────────────────────╮ │ Included: 18 fragments (14,820 tokens, 92.6% of budget) │ │ Excluded: 7 fragments (3,200 tokens, below budget cutoff) │ │ Deduplicated: 4 fragments merged │ │ Skeleton: 2,400 tokens (15% reserve) │ │ Preamble: 180 tokens │ ╰──────────────────────────────────────────────────────────────────╯
✓ OK Simulation complete (dry run — no context was assembled)
agents actor
Purpose Manage actors and run actor configurations directly.
agents actor run
agents actor run [(--output|-o) <OUTPUT_FILE>]
[-v...] [--unsafe|-u] [--context <CONTEXT_NAME>]
[--context-dir <CONTEXT_PATH>] [--load-context <LOAD_CONTEXT_NAME>]
[(--temperature|-t) <TEMP>] [--allow-rxpy-in-run-mode]
[--skill <SKILL>]... <NAME> <PROMPT>Purpose Run a named actor in isolation with simple, manual context.
Arguments
<NAME>: The name of the actor to run (required).<PROMPT>: Prompt text (positional argument).--output/-o FILE: Output file path.--verbose/-v: Increase log verbosity (repeatable). Same semantics as the global-vflag: no-v= normal output only,-v= ERROR,-vv= WARN,-vvv= INFO,-vvvv= DEBUG,-vvvvv= TRACE.--unsafe/-u: Allow unsafe configs.--context NAME: Named actor context to attach.--context-dir PATH: Context storage location.--load-context FILE: Load context from JSON.--temperature/-t FLOAT: Override temperature.--skill NAME: Skill to attach (repeatable).--allow-rxpy-in-run-mode: Allow RxPy routes in run mode.
Examples
$ agents actor run --context docs local/code_reader "Summarize the README"╭─ Run Summary ─────────────────╮ │ Actor: local/code_reader │ │ Context: docs │ │ Temperature: 0.2 │ │ Provider: anthropic │ │ Model: claude-3.5 │ ╰───────────────────────────────╯
╭─ Inputs ─────────────────────╮ │ Prompt: Summarize the README │ │ Context Files: 3 │ │ Context Size: 12.4 KB │ ╰──────────────────────────────╯
╭─ Result Metrics ───────╮ │ Output: stdout │ │ Input Tokens: 1,524 │ │ Output Tokens: 842 │ │ Duration: 1.8s │ │ Cost: $0.0021 │ │ Tool Calls: 0 │ ╰────────────────────────╯
✓ OK Summary generated
Running an actor with a custom temperature and skill attachment:
$ agents actor run --temperature 0.2 --skill local/code-analysis \ local/reviewer "Review the auth module for security issues"╭─ Actor Run ─────────────────────────╮ │ Actor: local/reviewer │ │ Type: agent │ │ Temperature: 0.2 │ │ Skill: local/code-analysis │ ╰─────────────────────────────────────╯
╭─ Response ────────────────────────────────────────────────────────╮ │ I identified 3 potential security concerns in the auth module: │ │ │ │ 1. Token expiry not validated in
validate_session()at │ │ src/auth/session.py:42. The JWT expiry claim is decoded but │ │ not checked against current time. │ │ │ │ 2. Missing rate limiting on the/auth/refreshendpoint. │ │ An attacker could brute-force refresh tokens. │ │ │ │ 3. Low severity: Password hashing uses bcrypt with cost=10. │ │ Consider cost=12 for production deployments. │ ╰───────────────────────────────────────────────────────────────────╯╭─ Usage ──────────────╮ │ Tokens: 2,840 │ │ Duration: 4.2s │ │ Cost: $0.009 │ │ Tool Calls: 6 │ ╰──────────────────────╯
✓ OK Actor run completed
Saving actor output to a file:
$ agents actor run --output review.md local/reviewer "Write a code review report"╭─ Actor Run ───────────────╮ │ Actor: local/reviewer │ │ Output: review.md │ │ Tokens: 4,120 │ │ Duration: 6.8s │ ╰───────────────────────────╯
✓ OK Output written to review.md (2.4 KB)
agents actor add
agents actor add (--config|-c) <FILE> [--update]Purpose
Add a new actor configuration, or replace an existing one with --update. The actor is fully defined by the YAML configuration file specified with --config. If a local actor with the same name already exists, the command fails unless the --update flag is provided, which replaces the existing actor.
Arguments
--config/-c FILE: Actor config file (required). The file must fully define the actor, including thenamefield which determines the actor's registered name.--update: Replace an existing local actor with the same name. Without this flag, attempting to add an actor whose name is already registered will fail.
Examples
$ agents actor add --config ./actors/reviewer.yaml╭─ Actor Added ────────╮ │ Name: local/reviewer │ │ Provider: openai │ │ Model: gpt-4 │ │ Default: yes │ │ Unsafe: no │ │ Type: graph │ ╰──────────────────────╯
╭─ Config ─────────────────────╮ │ Path: ./actors/reviewer.yaml │ │ Hash: 8b3f3d2 │ │ Options: 4 │ │ Nodes: 3 │ │ Edges: 4 │ ╰──────────────────────────────╯
╭─ Capabilities ───────╮ │ - code review │ │ - diff summarization │ │ - lint guidance │ ╰──────────────────────╯
╭─ Tools ────────────────────────╮ │ Tool Read-Only Safe │ │ ──────────── ───────── ──── │ │ read_file yes yes │ │ search_files yes yes │ │ git_diff yes yes │ ╰────────────────────────────────╯
✓ OK Actor added
Attempting to register an actor that already exists (without --update):
$ agents actor add --config ./actors/reviewer.yaml╭─ Error ─────────────────────────────────────────────────╮ │ Actor already exists: local/reviewer │ │ Registered: 2026-02-07 14:22 │ │ Use --update to replace the existing actor definition. │ ╰─────────────────────────────────────────────────────────╯
✗ ERROR Actor already registered — use --update to replace
Updating an existing actor with --update:
$ agents actor add --config ./actors/reviewer.yaml --update╭─ Actor Updated ─────────────────────╮ │ Name: local/reviewer │ │ Type: agent │ │ Status: updated │ ╰─────────────────────────────────────╯
╭─ Changes ────────────────────────────╮ │ Skills: +1 (local/code-analysis) │ │ Model: unchanged │ │ Graph: updated │ ╰──────────────────────────────────────╯
✓ OK Actor updated
agents actor remove
agents actor remove <NAME>Purpose Remove a custom actor.
Arguments
<NAME>: Actor name.
Examples
$ agents actor remove local/reviewer╭─ Actor Removed ──────╮ │ Name: local/reviewer │ │ Provider: openai │ │ Model: gpt-4 │ ╰──────────────────────╯
╭─ Impact ───────────────────────────────────╮ │ Sessions: 0 affected │ │ Active Plans: 0 affected │ │ Actions Referencing: 0 │ ╰────────────────────────────────────────────╯
╭─ Cleanup ──────────────╮ │ Config: kept on disk │ │ Contexts: 1 orphaned │ ╰────────────────────────╯
✓ OK Actor removed
agents actor list
agents actor listPurpose List all actors.
Arguments
None.
Examples
$ agents actor list╭─ Actors ──────────────────────────────────────────────────────────────╮ │ Name Provider Model Default Built-in Unsafe │ │ ────────────── ───────── ────────── ─────── ──────── ────── │ │ local/reviewer openai gpt-4 ✓ no │ │ openai/gpt-4 openai gpt-4 ✓ no │ │ anthropic/3.5 anthropic claude-3.5 ✓ no │ ╰───────────────────────────────────────────────────────────────────────╯
╭─ Summary ──────────────╮ │ Total: 3 │ │ Built-in: 2 │ │ Custom: 1 │ │ Unsafe: 0 │ │ Providers Used: 2 │ ╰────────────────────────╯
✓ OK 3 actors listed
agents actor show
agents actor show <NAME>Purpose Show details for a single actor.
Arguments
<NAME>: Actor name.
Examples
$ agents actor show local/reviewer╭─ Actor Details ────────────────────╮ │ Name: local/reviewer │ │ Provider: openai │ │ Model: gpt-4 │ │ Default: yes │ │ Built-in: no │ │ Unsafe: no │ │ Type: graph │ │ Created: 2026-02-08 12:35 │ │ Updated: 2026-02-08 12:40 │ │ Config: ./actors/reviewer.yaml │ │ Config Hash: 9c4e2a1 │ ╰────────────────────────────────────╯
╭─ Options ──────────╮ │ - temperature: 0.2 │ │ - max_tokens: 2048 │ │ - top_p: 1.0 │ ╰────────────────────╯
╭─ Graph Structure ─╮ │ Nodes: 3 │ │ Edges: 4 │ │ Entry: analyze │ │ Exit: report │ ╰───────────────────╯
╭─ Tools ────────────────────────╮ │ Tool Read-Only Safe │ │ ──────────── ───────── ──── │ │ read_file yes yes │ │ search_files yes yes │ │ git_diff yes yes │ ╰────────────────────────────────╯
╭─ Access ────────────╮ │ Unsafe: no │ │ Filesystem: allowed │ │ Network: restricted │ ╰─────────────────────╯
╭─ Usage ─────────────────────────────────────────╮ │ Referenced by Actions: 1 (local/code-coverage) │ │ Active in Sessions: 0 │ │ Total Runs: 14 │ │ Avg Cost/Run: $0.0032 │ ╰─────────────────────────────────────────────────╯
✓ OK Actor loaded
agents actor context
Purpose Manage manual context for actor runs. These commands mirror the legacy context behavior but are scoped to an actor context name.
agents actor context remove
agents actor context remove [--yes|-y] (--all|-a|<NAME>)Purpose Remove files or directories from an actor context.
Arguments
<NAME>: Context name (positional argument). Use --all/-a to target all contexts.--all, -a: Target all contexts instead of a named one.--yes, -y: Skip confirmation prompt.
Examples
$ agents actor context remove docs╭─ Context Removed ───────────╮ │ Context: docs │ │ Status: removed │ ╰─────────────────────────────╯
╭─ Stats ───────────────────╮ │ Remaining Size: 48 KB │ │ Updated: 2026-02-08 13:06 │ ╰───────────────────────────╯
✓ OK Context updated
agents actor context list
agents actor context list [<REGEX>]Purpose List files stored in an actor context.
Arguments
[REGEX]: Optional regex filter for context names.
Examples
$ agents actor context list docs╭─ Context Files ──────────────────────────────╮ │ Name Type Size Added │ │ ──────────────── ──── ─────── ────────── │ │ README.md file 4.2 KB 02-08 12:10 │ │ docs/overview.md file 12.8 KB 02-08 12:10 │ │ docs/cli.md file 9.5 KB 02-08 12:10 │ ╰──────────────────────────────────────────────╯
╭─ Stats ───────────────────────╮ │ Total Files: 3 │ │ Total Size: 26.5 KB │ │ Estimated Tokens: ~6,600 │ │ Languages: Markdown │ ╰───────────────────────────────╯
✓ OK 3 files listed
agents actor context show
agents actor context show <NAME>Purpose Show content of a file in an actor context.
Arguments
<NAME>: Context name (positional argument).
Examples
$ agents actor context show docs╭─ Context Summary ───────────╮ │ Context: docs │ │ Files: 3 │ │ Total Size: 26.5 KB │ │ Estimated Tokens: ~6,600 │ │ Created: 2026-02-08 12:10 │ ╰─────────────────────────────╯
✓ OK Context displayed
agents actor context export
agents actor context export (--output|-o) <FILE> <NAME>Purpose Export a context as JSON.
Arguments
<NAME>: Context name (positional argument).--output/-o FILE: Output file path (required).
Examples
$ agents actor context export --output /tmp/docs-context.json docs╭─ Context Export ───────────────╮ │ Context: docs │ │ Output: /tmp/docs-context.json │ │ Items: 12 │ │ Size: 48 KB │ ╰────────────────────────────────╯
╭─ Integrity ──────────────────╮ │ Checksum: sha256:19b2...a7d0 │ │ Compressed: no │ ╰──────────────────────────────╯
✓ OK Export completed
agents actor context import
agents actor context import [--update] (--input|-i) <FILE> [<NAME>]Purpose Import a context from JSON.
Arguments
[NAME]: Context name (optional, inferred from file if omitted).--input/-i FILE: Input JSON file (required).--update: Replace an existing context with the same name.
Examples
$ agents actor context import --input /tmp/docs-context.json docs╭─ Context Import ──────────────╮ │ Context: docs │ │ Input: /tmp/docs-context.json │ │ Items: 12 │ ╰───────────────────────────────╯
╭─ Merge ───────────╮ │ Strategy: replace │ │ Conflicts: 0 │ ╰───────────────────╯
✓ OK Import completed
agents actor context clear
agents actor context clear [--yes|-y] (--all|-a|<NAME>)Purpose Clear all files from a context but keep the context itself.
Arguments
<NAME>: Context name (positional argument). Use --all/-a to target all contexts.--all, -a: Target all contexts instead of a named one.--yes, -y: Skip confirmation.
Examples
$ agents actor context clear docsClear context docs? [y/N]: y
╭─ Context Cleared ────╮ │ Context: docs │ │ Items: 12 removed │ │ Storage: 48 KB freed │ ╰──────────────────────╯
╭─ Retention ────────╮ │ Context: preserved │ │ Files: removed │ ╰────────────────────╯
✓ OK Context cleared
agents skill
Purpose Manage skills — reusable, namespaced collections of tools. Skills are defined in their own YAML configuration files and registered in the system through these commands. Once registered, skills can be referenced by actors.
agents skill add
agents skill add (--config|-c) <FILE> [--update]Purpose
Register a new skill. The skill is fully defined by the YAML configuration file specified with --config. If a skill with the same name already exists, the command fails unless the --update flag is provided, which allows overwriting the existing registration with the new configuration.
Arguments
--config/-c FILE: Path to the skill YAML configuration file (required). The file must fully define the skill, including thenamefield which determines the skill's registered name.--update: Allow overwriting an existing skill registration. Without this flag, attempting to add a skill whose name is already registered will fail.
Examples
Registering a new skill:
$ agents skill add --config ./skills/devops-toolkit.yaml╭─ Skill Registered ────────────────────────╮ │ Name: local/devops-toolkit │ │ Description: Full-stack development tools │ │ Config: ./skills/devops-toolkit.yaml │ │ Created: 2026-02-08 13:10 │ ╰───────────────────────────────────────────╯
╭─ Includes ────────────────────╮ │ local/file-ops (registered) │ │ local/git-ops (registered) │ │ local/github (registered) │ ╰───────────────────────────────╯
╭─ Tool Sources ────────────────────────────────╮ │ Source Count Details │ │ ───────────── ───── ─────────────────── │ │ builtin 14 file, dir, git, shell │ │ mcp 6 github (4), linear (2) │ │ agent_skill 2 deploy, code-review │ │ custom 1 run_migrations │ │ ───────────── ───── ─────────────────── │ │ Total: 23 │ ╰───────────────────────────────────────────────╯
╭─ MCP Servers ────────────────────╮ │ linear: validated (2 tools) │ ╰──────────────────────────────────╯
✓ OK Skill registered with 23 tools
Attempting to add a skill that already exists (without --update):
$ agents skill add --config ./skills/devops-toolkit-v2.yaml✗ Error: Skill 'local/devops-toolkit' is already registered.
To overwrite the existing configuration, re-run with --update:
agents skill add --config ./skills/devops-toolkit-v2.yaml --update
Updating an existing skill with --update:
$ agents skill add --config ./skills/devops-toolkit-v2.yaml --update╭─ Skill Updated ───────────────────────────╮ │ Name: local/devops-toolkit │ │ Description: Full-stack development tools │ │ Updated: 2026-02-08 14:22 │ ╰───────────────────────────────────────────╯
╭─ Changes ─────────────────────╮ │ Tools Added: 2 │ │ Tools Removed: 0 │ │ Tools Modified: 1 │ │ Includes Changed: no │ │ MCP Servers Changed: no │ ╰───────────────────────────────╯
╭─ Affected Actors ─────────────────────────────╮ │ Warning: 2 actors reference this skill: │ │ - local/code-assistant │ │ - local/full-stack-assistant │ │ These actors will pick up changes on next use │ ╰───────────────────────────────────────────────╯
✓ OK Skill updated (23 → 25 tools)
agents skill remove
agents skill remove [--yes|-y] <NAME>Purpose Remove a registered skill.
Arguments
<NAME>: Skill name to remove.--yes: Skip confirmation prompt.
Examples
$ agents skill remove local/devops-toolkitRemove skill local/devops-toolkit? [y/N]: y
╭─ Skill Removed ──────────────────────╮ │ Name: local/devops-toolkit │ │ Tools: 23 removed from registry │ │ MCP Servers: 1 connection closed │ ╰──────────────────────────────────────╯
╭─ Dependency Check ─────────────────────────────────╮ │ Warning: 1 skill includes this skill: │ │ - local/full-stack-dev (will lose devops tools) │ │ Warning: 2 actors reference this skill: │ │ - local/code-assistant │ │ - local/full-stack-assistant │ ╰────────────────────────────────────────────────────╯
✓ OK Skill removed
agents skill list
agents skill list [(--namespace|-n) <NS>] [--source <SOURCE>]Purpose List registered skills with optional filters.
Arguments
--namespace/-n NS: Filter by namespace.--source SOURCE: Filter by tool source type (mcp, agent_skill, builtin, custom).
Examples
$ agents skill list╭─ Skills ────────────────────────────────────────────────────────────╮ │ Name Tools Includes Sources │ │ ────────────────────── ───── ──────── ────────────────────── │ │ local/file-ops 9 0 builtin │ │ local/git-ops 4 0 builtin │ │ local/github 4 0 mcp │ │ local/devops-toolkit 23 3 builtin, mcp, custom │ │ local/full-stack-dev 25 4 builtin, mcp, custom │ ╰─────────────────────────────────────────────────────────────────────╯
╭─ Summary ─────────╮ │ Total: 5 │ │ Local: 5 │ │ Server: 0 │ │ Total Tools: 28 │ ╰───────────────────╯
✓ OK 5 skills listed
agents skill show
agents skill show <NAME>Purpose Show full details for a registered skill, including its includes, tool sources, and capability summary.
Arguments
<NAME>: Skill name.
Examples
$ agents skill show local/devops-toolkit╭─ Skill Details ──────────────────────────────╮ │ Name: local/devops-toolkit │ │ Description: Full-stack development tools │ │ Config: ./skills/devops-toolkit.yaml │ │ Created: 2026-02-08 13:10 │ │ Updated: 2026-02-08 14:22 │ ╰──────────────────────────────────────────────╯
╭─ Includes (3) ────────────────────────────╮ │ local/file-ops → 9 tools (builtin) │ │ local/git-ops → 4 tools (builtin) │ │ local/github → 4 tools (mcp) │ ╰───────────────────────────────────────────╯
╭─ Direct Tools (6) ────────────────────────────────────────╮ │ Name Source Writes Checkpoint │ │ ──────────────── ─────────── ────── ────────── │ │ create_issue mcp:linear yes no │ │ list_issues mcp:linear no — │ │ deploy-to-staging agent_skill yes composite │ │ code-review agent_skill no — │ │ run_migrations custom yes transaction │ │ shell_execute builtin yes snapshot │ ╰───────────────────────────────────────────────────────────╯
╭─ MCP Servers (1) ───────────────────╮ │ linear: stdio, 2 tools, connected │ ╰─────────────────────────────────────╯
╭─ Capability Summary ──────────╮ │ Total Tools: 23 │ │ Read-Only: 10 │ │ Writes: 13 │ │ Checkpointable: 10 │ │ Has Side Effects: 3 │ │ Requires Approval: 1 │ ╰───────────────────────────────╯
╭─ Referenced By ───────────────────╮ │ Actors: local/code-assistant │ │ Skills: local/full-stack-dev │ ╰───────────────────────────────────╯
✓ OK Skill loaded
agents skill tools
agents skill tools <NAME>Purpose List all tools provided by a skill, including those inherited from included skills (the flattened tool set).
Arguments
<NAME>: Skill name.
Examples
$ agents skill tools local/devops-toolkit╭─ Tools for local/devops-toolkit ─────────────────────────────────────────────────────────╮ │ Tool Source From Skill Read-Only Writes Checkpoint │ │ ───────────────── ─────────── ────────────── ───────── ────── ────────── │ │ read_file builtin local/file-ops ✓ — — │ │ write_file builtin local/file-ops — ✓ file │ │ edit_file builtin local/file-ops — ✓ file │ │ delete_file builtin local/file-ops — ✓ file │ │ move_file builtin local/file-ops — ✓ file │ │ copy_file builtin local/file-ops — ✓ file │ │ create_directory builtin local/file-ops — ✓ file │ │ list_directory builtin local/file-ops ✓ — — │ │ delete_directory builtin local/file-ops — ✓ file │ │ git_status builtin local/git-ops ✓ — — │ │ git_diff builtin local/git-ops ✓ — — │ │ git_log builtin local/git-ops ✓ — — │ │ git_blame builtin local/git-ops ✓ — — │ │ create_issue mcp:github local/github — ✓ no │ │ create_pr mcp:github local/github — ✓ no │ │ list_repos mcp:github local/github ✓ — — │ │ get_file_contents mcp:github local/github ✓ — — │ │ create_issue mcp:linear (direct) — ✓ no │ │ list_issues mcp:linear (direct) ✓ — — │ │ run_migrations custom (direct) — ✓ transaction │ │ deploy-to-staging agent_skill (direct) — ✓ composite │ │ code-review agent_skill (direct) ✓ — — │ │ shell_execute builtin (direct) — ✓ snapshot │ ╰──────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Summary ──────────────────╮ │ Total: 23 │ │ From Includes: 17 │ │ Direct: 6 │ │ Read-Only: 10 │ │ Writes: 13 │ │ Checkpointable: 10 │ ╰────────────────────────────╯
✓ OK 23 tools listed
agents tool
Purpose Manage tools — namespaced, independently registered, callable operations. Tools are defined in their own YAML configuration files and registered in the system through these commands. Once registered, tools can be referenced by name in skills (as part of a tool collection) and in actor graphs (as tool nodes).
agents tool add
agents tool add (--config|-c) <FILE> [--update]Purpose
Register a new tool. The tool is fully defined by the YAML configuration file specified with --config. If a tool with the same name already exists, the command fails unless the --update flag is provided, which allows overwriting the existing registration with the new configuration.
Arguments
--config/-c FILE: Path to the tool YAML configuration file (required). The file must fully define the tool, including thenamefield which determines the tool's registered name.--update: Allow overwriting an existing tool registration. Without this flag, attempting to add a tool whose name is already registered will fail.
Examples
Registering a new tool:
$ agents tool add --config ./tools/run-migrations.yaml╭─ Tool Registered ────────────────────────────╮ │ Name: local/run-migrations │ │ Description: Run database migrations │ │ Source: custom │ │ Config: ./tools/run-migrations.yaml │ │ Created: 2026-02-09 10:15 │ ╰──────────────────────────────────────────────╯
╭─ Capability ─────────────────────╮ │ Writes: true │ │ Write Scope: database:migrations │ │ Checkpointable: true │ │ Checkpoint Scope: transaction │ │ Side Effects: schema_mutation │ ╰──────────────────────────────────╯
✓ OK Tool registered
Attempting to add a tool that already exists (without --update):
$ agents tool add --config ./tools/run-migrations-v2.yaml✗ Error: Tool 'local/run-migrations' is already registered.
To overwrite the existing configuration, re-run with --update:
agents tool add --config ./tools/run-migrations-v2.yaml --update
Updating an existing tool with --update:
$ agents tool add --config ./tools/db-migrate.yaml --update╭─ Tool Updated ─────────────────────────────╮ │ Name: local/db-migrate │ │ Source: custom (Python) │ │ Status: updated (was version 1 → now 2) │ ╰────────────────────────────────────────────╯
╭─ Changes ─────────────────────────────────────╮ │ Input Schema: modified (added batch_size) │ │ Capabilities: unchanged (writes, checkpoint) │ │ Description: updated │ ╰───────────────────────────────────────────────╯
╭─ References ──────────────────╮ │ Skills: 1 (local/db-tools) │ │ Actors: 0 │ │ Referencing skills will use │ │ the updated definition. │ ╰───────────────────────────────╯
✓ OK Tool updated
agents tool remove
agents tool remove [--yes|-y] <NAME>Purpose Remove a registered tool. Because Validations are a subtype of Tool and share the same registry, this command also removes validations. When removing a Validation, all attachments are automatically detached before the entry is removed from the registry.
Arguments
<NAME>: Tool or validation name to remove.--yes: Skip confirmation prompt.
Examples
$ agents tool remove local/run-migrationsRemove tool local/run-migrations? [y/N]: y
╭─ Tool Removed ────────────────────╮ │ Name: local/run-migrations │ │ Source: custom │ ╰───────────────────────────────────╯
╭─ References ──────────────────────────────────────╮ │ Warning: This tool is referenced by: │ │ - Skill: local/devops-toolkit │ │ - Actor graph node: code_executor.run_db_migrate │ │ These references will break until resolved. │ ╰───────────────────────────────────────────────────╯
✓ OK Tool removed
When removing a validation, attachments are cleaned up automatically:
$ agents tool remove local/check-bundle-sizeRemove validation local/check-bundle-size? [y/N]: y
╭─ Validation Removed ─────────────────────────────╮ │ Name: local/check-bundle-size │ │ Description: Check bundle size (advisory) │ ╰──────────────────────────────────────────────────╯
╭─ Detached From ────────────────────────────────────────────╮ │ Warning: This validation had 2 attachments: │ │ - local/api-repo (scope: project local/api-service) │ │ - local/api-repo (direct) │ │ These attachments have been removed. │ ╰───────────────────────────────────────────────────────────╯
✓ OK Validation removed
agents tool list
agents tool list [(--namespace|-n) <NS>] [--source <SOURCE>] [--type (tool|validation)] [<REGEX>]Purpose
List registered tools with optional filters. Because Validations are a subtype of Tool and share the same registry, this command lists both plain tools and validations. Use --type validation to show only validations, or --type tool to show only plain tools. When --type is omitted, both are listed with a Type column distinguishing them.
Arguments
--namespace/-n NS: Filter by namespace.--source SOURCE: Filter by tool source type (mcp, agent_skill, builtin, custom).--type TYPE: Filter by entry type:tool(plain tools only) orvalidation(validations only). When omitted, both are listed.
Examples
$ agents tool list --namespace local╭─ Tools ─────────────────────────────────────────────────────────────────╮ │ Name Type Source Read-Only Writes │ │ ───────────────────────── ────────── ─────── ───────── ────── │ │ local/run-migrations tool custom — ✓ │ │ local/validate-api-compat tool custom — ✓ │ │ local/create-subplan tool custom — ✓ │ │ local/deploy-staging tool agent — ✓ │ │ local/run-tests validation custom ✓ — │ │ local/lint-check validation custom ✓ — │ │ local/check-bundle-size validation custom ✓ — │ │ local/type-check validation custom ✓ — │ ╰─────────────────────────────────────────────────────────────────────────╯
╭─ Summary ────────────────╮ │ Total: 8 │ │ Tools: 4 │ │ Validations: 4 │ │ Read-Only: 4 │ │ Writes: 4 │ ╰──────────────────────────╯
✓ OK 8 tools listed
$ agents tool list --type validation --namespace local╭─ Validations ────────────────────────────────────────────────────────────────────╮ │ Name Source Mode Attachments │ │ ───────────────────────── ─────── ──────────── ──────────────────────── │ │ local/run-tests custom required 2 (1 direct, 1 project) │ │ local/lint-check custom required 1 (1 direct) │ │ local/check-bundle-size custom informational 1 (1 project) │ │ local/type-check custom required 2 (2 project) │ ╰──────────────────────────────────────────────────────────────────────────────────╯
╭─ Summary ──────────────╮ │ Total: 4 │ │ Required: 3 │ │ Informational: 1 │ ╰────────────────────────╯
✓ OK 4 validations listed
agents tool show
agents tool show <NAME>Purpose
Show full details for a registered tool, including its schema, capability metadata, and where it is referenced. Because Validations are a subtype of Tool and share the same registry, this command also shows validation details. When showing a Validation, additional validation-specific fields are displayed: Mode (required/informational) and Attached To (listing each resource attachment with its scope — direct, project-scoped, or plan-scoped).
Arguments
<NAME>: Tool or validation name.
Examples
$ agents tool show local/run-migrations╭─ Tool Details ───────────────────────────────╮ │ Name: local/run-migrations │ │ Description: Run database migrations │ │ Source: custom │ │ Config: ./tools/run-migrations.yaml │ │ Registered: 2026-02-09 10:15 │ ╰──────────────────────────────────────────────╯
╭─ Input Schema ─────────────────────────────────╮ │ direction: string (required) enum: [up, down] │ │ count: integer (default: 1) │ ╰────────────────────────────────────────────────╯
╭─ Capability ──────────────────────╮ │ Read-Only: false │ │ Writes: true │ │ Write Scope: database:migrations │ │ Checkpointable: true │ │ Checkpoint Scope: transaction │ │ Side Effects: schema_mutation │ │ Idempotent: false │ ╰───────────────────────────────────╯
╭─ Referenced By ───────────────────────╮ │ Skills: │ │ - local/devops-toolkit │ │ Actor Graph Nodes: │ │ - code_executor.run_db_migrate │ ╰───────────────────────────────────────╯
✓ OK Tool details loaded
When showing a Validation, the output includes validation-specific sections:
$ agents tool show local/run-tests╭─ Validation Details ──────────────────────────────╮ │ Name: local/run-tests │ │ Type: validation │ │ Description: Run unit tests with coverage │ │ Source: custom │ │ Mode: required │ │ Config: ./validations/run-tests.yaml │ │ Registered: 2026-02-09 10:15 │ ╰───────────────────────────────────────────────────╯
╭─ Input Schema ─────────────────────────────────────╮ │ coverage_threshold: integer (default: 80) │ ╰────────────────────────────────────────────────────╯
╭─ Output Schema (Validation Return Format) ─────────╮ │ passed: boolean (required) │ │ message: string (optional) │ │ data: object (optional, arbitrary structure) │ ╰────────────────────────────────────────────────────╯
╭─ Capability ──────────────────────╮ │ Read-Only: true (enforced) │ │ Checkpointable: false (enforced) │ │ Timeout: 600s │ ╰───────────────────────────────────╯
╭─ Attached To ─────────────────────────────────────────────────────╮ │ local/api-repo (direct, always active) │ │ (attachment: 01HXM5B2C3D4E5F6G7H8J9K0L1) │ │ local/api-repo (scope: project local/api-service) │ │ (attachment: 01HXM5A1B2C3D4E5F6G7H8J9K0) │ ╰──────────────────────────────────────────────────────────────────╯
✓ OK Validation details loaded
agents validation
Purpose
Register and attach validations — a specialized subtype of Tool that extends the Tool class with pass/fail semantics and attachment scoping. A Validation inherits all base Tool properties (name, source, input_schema, capability, resource bindings, lifecycle hooks, etc.) and adds: a mode field (required or informational), and a structured JSON return format with a mandatory passed boolean, an optional message string, and an optional data object for arbitrary informational output. Since Validation extends Tool, it shares the same backing sources (custom, MCP, agent_skill, builtin) plus the additional wrapped source type for Validations that wrap an existing Tool via the wraps field. It uses the same YAML configuration format (with an additional validation block and optional wraps/transform fields), and the same registries and resolution mechanisms. Validations are always read-only (writes is always false, checkpointable is always false).
Because Validations are a subtype of Tool and share the same Tool Registry and naming namespace, only validation-specific operations have dedicated subcommands here: add (registration with validation-specific fields like mode), attach, and detach. General tool operations — listing, inspection, and removal — are handled through the agents tool commands:
- List validations:
agents tool list --type validation - Show validation details:
agents tool show <NAME>(displays validation-specific fields when the entry is a Validation) - Remove a validation:
agents tool remove <NAME>(automatically removes all attachments)
A validation is always attached to a resource. The optional --project or --plan flag controls the scope of the attachment:
- Direct attachment (no scope flag): The validation is always active for that resource, regardless of which plan or project accesses it. Use
agents validation attach <RESOURCE> <VALIDATION> [args...]. - Project-scoped attachment (
--project): The validation only runs for that resource when it is being interacted with through the specified project. Useagents validation attach --project <PROJECT> <RESOURCE> <VALIDATION> [args...]. - Plan-scoped attachment (
--plan): The validation only runs for that resource when it is being interacted with through the specified plan. Useagents validation attach --plan <PLAN_ID> <RESOURCE> <VALIDATION> [args...].
The same validation can be attached to the same resource multiple times — directly, through different projects, and through different plans — and since validations accept arguments, attaching the same validation to the same resource multiple times with different arguments is a common and useful configuration pattern (e.g., running the same test suite with different coverage thresholds for different projects).
Each attach invocation returns a system-assigned attachment ULID uniquely identifying that attachment. Use this ULID with agents validation detach <ATTACHMENT_ID> to remove a specific attachment.
When multiple attachment scopes apply, the union of all applicable validations is used. A resource with a directly-attached validation will always have that validation run, plus any additional validations attached through the active project or plan.
agents validation add
agents validation add (--config|-c) <FILE> [--update]Purpose
Register a new validation. The validation is fully defined by the YAML configuration file specified with --config. The validation is registered in the shared Tool Registry under the name specified in the config file; the name must not conflict with any existing tool or validation. If a validation with the same name already exists, the command fails unless --update is provided. The validation YAML may use wraps to reference an existing Tool, reusing its implementation and interpreting its output through a transform function (see Tool Wrapping under Core Concepts > Validation). When wraps is used, the wrapped Tool must already be registered.
Arguments
--config/-c FILE: Path to the validation YAML configuration file (required). The file must fully define the validation, including thenamefield which determines the validation's registered name, thevalidation.modefield (requiredorinformational), and either inline implementation (source+code) or wrapping (wraps+transform).--update: Allow overwriting an existing validation registration.
Examples
$ agents validation add --config ./validations/run-tests.yaml╭─ Validation Registered ────────────────────────────────╮ │ Name: local/run-tests │ │ Description: Run unit tests with coverage │ │ Source: custom │ │ Mode: required │ │ Config: ./validations/run-tests.yaml │ │ Created: 2026-02-09 10:15 │ ╰────────────────────────────────────────────────────────╯
╭─ Capability ────────────────────────╮ │ Read-Only: true (enforced) │ │ Checkpointable: false (enforced) │ │ Timeout: 600s │ ╰─────────────────────────────────────╯
✓ OK Validation registered
$ agents validation add --config ./validations/lint-check.yaml╭─ Validation Registered ──────────────╮ │ Name: local/lint-check │ │ Description: Lint check │ │ Source: custom │ │ Mode: required │ │ Timeout: 300s │ ╰───────────────────────────────────────╯
✓ OK Validation registered
$ agents validation add --config ./validations/check-bundle-size.yaml╭─ Validation Registered ───────────────────────────╮ │ Name: local/check-bundle-size │ │ Description: Check bundle size (advisory) │ │ Source: custom │ │ Mode: informational │ │ Timeout: 300s │ ╰───────────────────────────────────────────────────╯
✓ OK Validation registered
agents validation attach
agents validation attach [--project <PROJECT>|--plan <PLAN_ID>]
<RESOURCE> <VALIDATION> [<ARGS>...]Purpose
Attach a registered validation to a resource, with an optional project or plan scope. A resource is always required — validations are fundamentally resource-centric. The optional --project or --plan flag narrows when the validation is active: without either flag, the validation runs whenever any plan or project accesses the resource; with --project, it only runs when the resource is accessed through that project; with --plan, it only runs when the resource is accessed through that plan. At most one scope flag may be provided per invocation.
The command returns a system-assigned attachment ULID that uniquely identifies this specific attachment association. This ULID is used to detach the validation later via agents validation detach <ATTACHMENT_ID>. The same validation can be attached to the same resource multiple times — directly, through different projects, through different plans, and even to the same resource and scope multiple times with different arguments.
Arguments
<RESOURCE>: Resource name or ULID to attach the validation to (positional argument, required).<VALIDATION>: Validation name (positional argument, required).--project PROJECT: Scope the attachment to a specific project. The validation only runs for this resource when it is accessed through the specified project.--plan PLAN_ID: Scope the attachment to a specific plan. The validation only runs for this resource when it is accessed through the specified plan.<ARGS>...: Optional validation-specific arguments (passed through to the validation tool'sinput_schemaat execution time).
Examples
$ agents validation attach --project local/api-service local/api-repo local/run-tests╭─ Validation Attached ──────────────────────────────────────╮ │ Attachment ID: 01HXM5A1B2C3D4E5F6G7H8J9K0 │ │ Validation: local/run-tests │ │ Mode: required │ │ Resource: local/api-repo │ │ Scope: project local/api-service │ ╰────────────────────────────────────────────────────────────╯
✓ OK Validation attached
$ agents validation attach local/api-repo local/lint-check╭─ Validation Attached ──────────────────────────────────────╮ │ Attachment ID: 01HXM5B2C3D4E5F6G7H8J9K0L1 │ │ Validation: local/lint-check │ │ Mode: required │ │ Resource: local/api-repo │ │ Scope: direct (always active) │ │ This validation will run for ALL plans/projects │ │ that access this resource. │ ╰────────────────────────────────────────────────────────────╯
✓ OK Validation attached
Attaching with validation-specific arguments:
$ agents validation attach --project local/api-service local/api-repo local/run-tests --coverage-threshold 90╭─ Validation Attached ──────────────────────────────────────╮ │ Attachment ID: 01HXM5C3D4E5F6G7H8J9K0L1M2 │ │ Validation: local/run-tests │ │ Mode: required │ │ Resource: local/api-repo │ │ Scope: project local/api-service │ │ Args: coverage_threshold=90 │ ╰────────────────────────────────────────────────────────────╯
✓ OK Validation attached
agents validation detach
agents validation detach [--yes|-y] <ATTACHMENT_ID>Purpose
Detach a validation by its attachment ULID. The attachment ULID is returned by agents validation attach when the attachment is created and is also displayed by agents tool show <VALIDATION> in the "Attached To" section. Because a validation may be attached to the same resource multiple times — directly, through different projects, through different plans, or even to the same resource and scope with different arguments — the ULID uniquely identifies the specific attachment to remove. The validation itself remains registered in the Tool Registry; only the attachment association is removed.
Arguments
<ATTACHMENT_ID>: The ULID of the attachment to remove (returned byagents validation attach).--yes, -y: Skip confirmation prompt.
Examples
$ agents validation detach 01HXM5A1B2C3D4E5F6G7H8J9K0Detach validation local/run-tests from resource local/api-repo (scope: project local/api-service)? [y/N]: y
╭─ Validation Detached ───────────────────────────────────────╮ │ Attachment ID: 01HXM5A1B2C3D4E5F6G7H8J9K0 │ │ Validation: local/run-tests │ │ Resource: local/api-repo │ │ Scope: project local/api-service │ ╰─────────────────────────────────────────────────────────────╯
✓ OK Validation detached
agents resource type
Purpose
Manage resource type definitions. Resource types are schema-level definitions that constrain categories of resources — they define what CLI arguments agents resource add <type> accepts, whether instances are physical or virtual, allowed parent/child types, auto-discovery behavior, sandbox strategy, and handler implementation. Built-in types (git-checkout, git, fs-mount, fs-directory, and their child types) are hardcoded. Custom types are defined in YAML configuration files and extend the system with new agents resource add subcommands.
agents resource type add
agents resource type add (--config|-c) <FILE> [--update]Purpose
Register a new custom resource type. The resource type is fully defined by the YAML configuration file specified with --config.
Arguments
--config/-c FILE: Path to the resource type YAML configuration file (required). The file must fully define the resource type, including thenamefield which determines the type's registered name.--update: If the type name already exists, overwrite it. Without this flag, adding a duplicate name fails with an error.
Examples
$ agents resource type add --config ./resource-types/svn.yaml╭─ Resource Type ────────────────────╮ │ Name: local/svn │ │ Physical/Virtual: physical │ │ User Addable: yes │ │ Registered: 2026-02-09 10:15 │ ╰────────────────────────────────────╯
╭─ CLI Arguments ────────────────────────╮ │ Argument Required Description │ │ ──────────── ──────── ───────────── │ │ --url yes Repository URL │ │ --checkout no Local checkout │ ╰────────────────────────────────────────╯
╭─ Child Types ──────────────────────────╮ │ Auto-discover: svn-revision, svn-file │ │ Manual link: fs-mount │ ╰────────────────────────────────────────╯
╭─ Sandbox ──────────────────────╮ │ Strategy: copy_on_write │ │ Handler: SVNHandler (custom) │ ╰────────────────────────────────╯
✓ OK Resource type registered ℹ New subcommand available: agents resource add local/svn
agents resource type remove
agents resource type remove [--yes|-y] <NAME>Purpose Remove a custom resource type. Built-in types cannot be removed. Fails if any registered resources use this type.
Arguments
<NAME>: Resource type name.--yes: Skip confirmation prompt.
Examples
$ agents resource type remove local/svnRemove resource type local/svn? [y/N]: y
╭─ Resource Type Removed ───────╮ │ Name: local/svn │ │ Resources Using: 0 │ │ Subcommand Removed: yes │ ╰───────────────────────────────╯
✓ OK Resource type removed
agents resource type list
agents resource type list [<REGEX>]Purpose List all registered resource types (built-in and custom).
Arguments
None (use the global --format option to control output format).
Examples
$ agents resource type list╭─ Resource Types ──────────────────────────────────────────────────────────────────────────────────╮ │ Name Source Phys/Virt Addable Auto-children │ │ ─────────────────── ──────── ───────── ─────── ────────────────────────────────────── │ │ git-checkout built-in physical yes git, fs-directory │ │ git built-in physical yes git-remote, git-branch, git-tag, │ │ git-commit, git-stash, git-submodule │ │ git-remote built-in physical no (none) │ │ git-branch built-in physical no git-commit │ │ git-tag built-in physical no (none) │ │ git-commit built-in physical no git-tree │ │ git-tree built-in physical no git-tree-entry │ │ git-tree-entry built-in physical no (none) │ │ git-stash built-in physical no (none) │ │ git-submodule built-in physical no (none) │ │ fs-mount built-in physical yes fs-directory │ │ fs-directory built-in physical yes fs-file, fs-directory, │ │ fs-symlink, fs-hardlink │ │ fs-file built-in physical no (none) │ │ fs-symlink built-in physical no (none) │ │ fs-hardlink built-in physical no (none) │ │ file built-in virtual no (none) │ │ directory built-in virtual no (none) │ │ symlink built-in virtual no (none) │ │ commit built-in virtual no (none) │ │ branch built-in virtual no (none) │ │ tag built-in virtual no (none) │ │ remote built-in virtual no (none) │ │ submodule built-in virtual no (none) │ │ tree built-in virtual no (none) │ │ local/svn custom physical yes svn-revision, svn-file │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Summary ─────────────────╮ │ Built-in: 24 │ │ Custom: 1 │ │ User Addable: 4 │ ╰───────────────────────────╯
✓ OK 25 resource types listed
agents resource type show
agents resource type show <NAME>Purpose Show detailed information about a resource type, including its full schema.
Arguments
<NAME>: Resource type name.
Examples
$ agents resource type show git-checkout╭─ Resource Type ──────────────────────────────────────────────────────╮ │ Name: git-checkout │ │ Source: built-in │ │ Description: A locally checked-out git repository with worktree │ │ Physical/Virtual: physical │ │ User Addable: yes │ ╰──────────────────────────────────────────────────────────────────────╯
╭─ CLI Arguments (agents resource add git-checkout) ─────────────────╮ │ Argument Required Type Description │ │ ────────── ──────── ────── ────────────────────────────── │ │ --path yes path Local checkout directory │ │ --branch no string Default branch (default: main) │ ╰────────────────────────────────────────────────────────────────────╯
╭─ Parent Types ───────────────╮ │ Allowed: (any, top-level OK) │ ╰──────────────────────────────╯
╭─ Child Types ────────────────────────────────────────────────────╮ │ Type Auto Manual Link Description │ │ ───────────── ──── ─────────── ────────────────────────── │ │ git yes no Repo object DB and history │ │ fs-directory yes no Worktree root directory │ ╰──────────────────────────────────────────────────────────────────╯
╭─ Sandbox ──────────────────────╮ │ Strategy: git_worktree │ │ Checkpointable: yes │ │ Handler: GitCheckoutHandler │ ╰────────────────────────────────╯
✓ OK Resource type details loaded
agents resource
Purpose
Manage independently registered resources. Resources represent anything that can be read, written, or queried — git repositories, filesystems, databases, APIs, and more. Resources are registered independently of projects and linked to one or more projects via agents project link-resource. Resources form a directed acyclic graph (DAG) with parent/child relationships, and child resources are often auto-discovered when a parent resource is registered.
agents resource add
agents resource add [(--description|-d) <DESC>] [--update] <TYPE> <NAME> [type-specific-flags...]Purpose
Register a new resource. Every resource receives a system-assigned ULID. User-added resources also require a namespaced name ([[server:]namespace/]name) specified at creation time. The add command uses type-specific subcommands — each registered resource type (built-in or custom) provides its own subcommand with type-appropriate arguments. On success, the command returns both the name and the ULID assigned to the newly created resource.
Arguments
<TYPE>: Resource type name (e.g.,git-checkout,git,fs-mount,fs-directory,local/svn). Determines type-specific flags.<NAME>: Namespaced resource name (positional, required). Must be globally unique following the standard[[server:]namespace/]nameformat.--description/-d TEXT: Optional resource description.
Type-specific flags depend on the resource type. See agents resource type show <TYPE> for available arguments.
The command returns the name and ULID of the newly created resource. Auto-discovered child resources are created automatically with ULIDs only (no names).
Examples
$ agents resource add git-checkout local/api-repo --path /home/user/projects/api-service --branch main╭─ Resource ──────────────────────────────────────╮ │ Name: local/api-repo │ │ ID: 01HXR1A1B2C3D4E5F6G7H8J9K0 │ │ Type: git-checkout │ │ Physical/Virtual: physical │ │ Path: /home/user/projects/api-service │ │ Branch: main │ │ Created: 2026-02-09 10:20 │ ╰─────────────────────────────────────────────────╯
╭─ Auto-discovered Children ─────────────────────────────────────────╮ │ ID Type Status │ │ ─────────────── ────────────── ───────────────── │ │ 01HXR1A1B2C3… git created │ │ 01HXR1A1B2C4… git-remote created │ │ 01HXR1A1B2C5… git-branch created │ │ 01HXR1A1B2C6… git-branch created │ │ 01HXR1A1B2C7… fs-directory created │ │ + 47 git-commit resources │ │ + 312 git-tree-entry resources │ │ + 3 fs-directory + 28 fs-file │ ╰────────────────────────────────────────────────────────────────────╯
╭─ Capabilities ─────────────────╮ │ Readable: yes │ │ Writable: yes │ │ Sandboxable: yes │ │ Checkpointable: yes │ │ Sandbox Strategy: git_worktree │ ╰────────────────────────────────╯
✓ OK Resource registered (395 child resources discovered)
$ agents resource add fs-mount local/docs --mount-path /docs/api-reference╭─ Resource ──────────────────────────────────────╮ │ Name: local/docs │ │ ID: 01HXR2B3C4D5E6F7G8H9J0K1L2 │ │ Type: fs-mount │ │ Physical/Virtual: physical │ │ Mount Path: /docs/api-reference │ │ Created: 2026-02-09 10:22 │ ╰─────────────────────────────────────────────────╯
╭─ Auto-discovered Children ────────────────────╮ │ + 1 fs-directory (root) │ │ + 3 fs-directory resources │ │ + 28 fs-file resources │ ╰───────────────────────────────────────────────╯
╭─ Capabilities ──────────────────────╮ │ Readable: yes │ │ Writable: yes │ │ Sandboxable: yes │ │ Checkpointable: yes │ │ Sandbox Strategy: copy_on_write │ ╰─────────────────────────────────────╯
✓ OK Resource registered (31 child resources discovered)
agents resource remove
agents resource remove [--yes|-y] <NAME>Purpose Remove a registered resource and all its auto-discovered child resources. Only user-added (named) resources can be removed. Fails if the resource is linked to any project (unlink first).
Arguments
<NAME>: Resource name (only user-added resources can be removed).--yes: Skip confirmation prompt.
Examples
$ agents resource remove local/api-repoRemove resource local/api-repo and 395 child resources? [y/N]: y
╭─ Resource Removed ──────────────────────────────────╮ │ Name: local/api-repo │ │ Type: git-checkout │ │ Children Removed: 395 │ │ Projects Unlinked: 0 │ ╰─────────────────────────────────────────────────────╯
✓ OK Resource removed
agents resource list
agents resource list [--all] [(--type|-t) <TYPE>] Purpose
List registered resources with optional filters. By default, lists only user-added (named) resources. Use --all to include auto-discovered child resources.
Arguments
--type/-t TYPE: Filter by resource type (accepts a regex).--all: Include auto-discovered child resources (by default only user-added resources are shown).
Examples
$ agents resource list╭─ Resources ──────────────────────────────────────────────────────────────────────────────────────────────╮ │ Name ID Type Phys/Virt Children Projects │ │ ───────────────── ──────────────────────────── ─────────── ───────── ──────── ──────────────── │ │ local/api-repo 01HXR1A1B2C3D4E5F6G7H8J9K0 git-checkout physical 395 local/api-service │ │ local/docs 01HXR2B2C3D4E5F6G7H8J9K0L1 fs-mount physical 32 local/api-service │ │ local/staging-db 01HXR3C3D4E5F6G7H8J9K0L1M2 database physical 12 local/api-service, │ │ local/staging │ ╰──────────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Summary ───────────╮ │ Total: 3 │ │ Physical: 3 │ │ Virtual: 0 │ │ Total Children: 405 │ ╰─────────────────────╯
✓ OK 3 resources listed
agents resource show
agents resource show <RESOURCE>Purpose Show detailed information about a registered resource including its type, capabilities, parent/child relationships, and linked projects.
Arguments
<RESOURCE>: Resource name or ULID.
Examples
$ agents resource show local/api-repo╭─ Resource ─────────────────────────────────────────╮ │ Name: local/api-repo │ │ ID: 01HXR1A1B2C3D4E5F6G7H8J9K0 │ │ Type: git-checkout │ │ Physical/Virtual: physical │ │ Path: /home/user/projects/api-service │ │ Branch: main │ │ Created: 2026-02-09 10:20 │ ╰────────────────────────────────────────────────────╯
╭─ Capabilities ────────────────────╮ │ Readable: yes │ │ Writable: yes │ │ Sandboxable: yes │ │ Checkpointable: yes │ │ Sandbox Strategy: git_worktree │ ╰───────────────────────────────────╯
╭─ Parents ────╮ │ (top-level) │ ╰──────────────╯
╭─ Direct Children ──────────────────────────────────────────────────────╮ │ ID Type Auto Children │ │ ──────────────────────────── ──────────── ──── ──────── │ │ 01HXR4D4E5F6G7H8J9K0L1M2N3 git yes 363 │ │ 01HXR5E5F6G7H8J9K0L1M2N3P4 fs-directory yes 32 │ ╰────────────────────────────────────────────────────────────────────────╯
╭─ Linked Projects ─────────────╮ │ - local/api-service (r/w) │ ╰───────────────────────────────╯
╭─ Tool Bindings ────────────────────────────────╮ │ Tool Slot Access │ │ ──────────────────────── ───── ─────────── │ │ (builtin) git_status repo read_only │ │ (builtin) git_diff repo read_only │ │ (builtin) git_log repo read_only │ │ local/run-migrations db (not bound) │ ╰────────────────────────────────────────────────╯
✓ OK Resource details loaded
agents resource inspect
agents resource inspect [--tree] [--file <PATH>] <RESOURCE>Purpose Inspect a resource's contents. For file-based resources (git checkouts, filesystem mounts, directories), displays the file tree or the contents of a specific file. Works with any resource — user-added or auto-discovered — via name or ULID.
Arguments
<RESOURCE>: Resource name or ULID.--tree: Show the resource's file/content tree.--file PATH: Show contents of a specific file within the resource.
Examples
$ agents resource inspect local/api-repo --tree╭─ Resource Tree: local/api-repo ──────────────────────────────────────╮ │ │ │ /home/user/projects/api-service │ │ ├── src/ │ │ │ ├── api/ │ │ │ │ ├── __init__.py │ │ │ │ ├── auth/ │ │ │ │ │ ├── auth_middleware.py │ │ │ │ │ ├── token_service.py │ │ │ │ │ └── rbac.py │ │ │ │ ├── routes/ │ │ │ │ └── models/ │ │ │ └── utils/ │ │ ├── tests/ │ │ ├── README.md │ │ └── pyproject.toml │ │ │ ╰─────────────────────────────────────────────────────────────────────╯
╭─ Summary ──────────────╮ │ Directories: 8 │ │ Files: 41 │ │ Total Size: 284 KB │ ╰────────────────────────╯
✓ OK Resource tree displayed
$ agents resource inspect local/api-repo --file src/api/auth/auth_middleware.py╭─ File: src/api/auth/auth_middleware.py ──────────────────╮ │ Resource: local/api-repo │ │ Size: 2.4 KB │ │ Language: Python │ │ Last Modified: 2026-02-10 14:32 │ ╰──────────────────────────────────────────────────────────╯
1 │ from fastapi import Request, HTTPException 2 │ from .token_service import validate_token 3 │ 4 │ async def auth_middleware(request: Request): 5 │ token = request.headers.get("Authorization") 6 │ if not token: 7 │ raise HTTPException(status_code=401) 8 │ user = await validate_token(token) 9 │ request.state.user = user 10 │ return user │ ... (24 more lines)
✓ OK File displayed
agents resource tree
agents resource tree [(--depth|-d) <N>] [(--type|-t) <TYPE>] <RESOURCE>Purpose Display the resource DAG as a tree starting from a given resource, showing parent/child relationships.
Arguments
<RESOURCE>: Resource name or ULID.--depth/-d N: Maximum depth to display (default: 3).--type/-t TYPE: Filter to only show children of a specific type.
Examples
$ agents resource tree local/api-repo --depth 2╭─ Resource Tree: local/api-repo ──────────────────────────────────────╮ │ │ │ git-checkout 01HXR1A1..J9K0 (physical) │ │ ├── git 01HXR4D4..M2N3 (physical) │ │ │ ├── git-remote 01HXR6F6..P4Q5 │ │ │ ├── git-branch 01HXR7G7..Q5R6 │ │ │ │ ├── git-commit 01HXR8H8..R6S7 │ │ │ │ └── ... 22 more git-commit resources │ │ │ └── git-branch 01HXR9J9..S7T8 │ │ │ └── ... 26 git-commit resources │ │ └── fs-directory 01HXR5E5..N3P4 (physical) │ │ ├── fs-directory 01HXRAK1..T8U9 │ │ ├── fs-directory 01HXRBL2..U9V0 │ │ └── ... 28 more fs-file resources │ │ │ ╰─────────────────────────────────────────────────────────────────────╯
╭─ Summary ──────────────╮ │ Total shown: 14 │ │ Total in subtree: 395 │ │ Max depth: 2 │ ╰────────────────────────╯
✓ OK Resource tree displayed
agents resource link-child
agents resource link-child <PARENT> <CHILD>Purpose Manually link one resource as a child of another. The child resource must already be registered. The resource types must be compatible (the parent type must allow the child type). A resource can have multiple parents.
Arguments
<PARENT>: Parent resource name or ULID.<CHILD>: Child resource name or ULID.
Examples
$ agents resource link-child local/api-repo 01HXR2B2C3D4E5F6G7H8J9K0L1╭─ Child Linked ──────────────────────────────────────╮ │ Parent: local/api-repo │ │ Child: 01HXR2B2C3D4E5F6G7H8J9K0L1 │ │ Child Type: fs-mount │ │ Status: linked │ ╰─────────────────────────────────────────────────────╯
✓ OK Child resource linked
agents resource unlink-child
agents resource unlink-child [--yes|-y] <PARENT> <CHILD>Purpose Remove a manual parent/child link between two resources. Auto-discovered links cannot be manually unlinked.
Arguments
<PARENT>: Parent resource name or ULID.<CHILD>: Child resource name or ULID.--yes: Skip confirmation prompt.
Examples
$ agents resource unlink-child local/api-repo 01HXR2B2C3D4E5F6G7H8J9K0L1Unlink 01HXR2B2C3D4E5F6G7H8J9K0L1 from parent local/api-repo? [y/N]: y
╭─ Child Unlinked ────────────────────────────────────╮ │ Parent: local/api-repo │ │ Child: 01HXR2B2C3D4E5F6G7H8J9K0L1 │ │ Status: unlinked │ ╰─────────────────────────────────────────────────────╯
✓ OK Child resource unlinked
agents plan
Purpose Manage plans through the Strategize -> Execute -> Apply lifecycle (instantiated from Action templates).
agents plan list
agents plan list [--phase <PHASE>] [--state <STATE>] [--project <PROJECT>]
[--action <ACTION>] [<REGEX>]Purpose List plans with optional filtering.
Arguments
--phase PHASE: Filter by phase.--state STATE: Filter by processing state.--project PROJECT: Filter by project.--action ACTION: Filter by action name.
Examples
$ agents --format table plan list --phase execute╭─ Plans ──────────────────────────────────────────────────────────────────────────────╮ │ ID Phase State Action Project Elapsed │ │ ──────── ─────── ────────── ─────────────────── ───────────────── ───────── │ │ 01HXM7A9 execute processing local/code-coverage local/api-service 00:01:12 │ ╰──────────────────────────────────────────────────────────────────────────────────────╯
╭─ Filters ──────╮ │ Phase: execute │ │ State: (any) │ │ Project: (any) │ │ Action: (any) │ ╰────────────────╯
╭─ Summary ─────────╮ │ Total: 1 │ │ Processing: 1 │ │ Completed: 0 │ │ Errored: 0 │ ╰───────────────────╯
✓ OK 1 plan listed
Listing all plans without filters shows plans across all phases and states:
$ agents plan list╭─ Plans ──────────────────────────────────────────────────────────────────────────────────────╮ │ ID Phase State Action Project Elapsed │ │ ──────── ────────── ────────── ─────────────────── ───────────────── ───────── │ │ 01HXM7A9 execute processing local/code-coverage local/api-service 00:01:12 │ │ 01HXM6R3 apply applied local/add-auth local/api-service 00:07:14 │ │ 01HXM5K2 execute errored local/migrate-db local/api-service 00:04:33 │ │ 01HXM4J1 strategize processing local/refactor-api local/web-app 00:00:45 │ │ 01HXM3H8 cancelled cancelled local/add-logging local/api-service 00:02:10 │ ╰──────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Summary ─────────────╮ │ Total: 5 │ │ Processing: 2 │ │ Completed: 1 │ │ Errored: 1 │ │ Cancelled: 1 │ ╰───────────────────────╯
✓ OK 5 plans listed
Filtering by project with --project:
$ agents plan list --project local/web-app╭─ Plans ──────────────────────────────────────────────────────────────────────────────╮ │ ID Phase State Action Project Elapsed │ │ ──────── ────────── ────────── ─────────────────── ───────────── ───────── │ │ 01HXM4J1 strategize processing local/refactor-api local/web-app 00:00:45 │ ╰──────────────────────────────────────────────────────────────────────────────────────╯
╭─ Filters ─────────────────╮ │ Phase: (any) │ │ State: (any) │ │ Project: local/web-app │ │ Action: (any) │ ╰───────────────────────────╯
✓ OK 1 plan listed
agents plan use
agents plan use [--automation-profile <PROFILE>]
[--invariant <INVARIANT>]...
[--strategy-actor <STRATEGY_ACTOR>]
[--execution-actor <EXEC_ACTOR>]
[--estimation-actor <EST_ACTOR>]
[--invariant-actor <INV_ACTOR>]
[--arg/-a name=value]...
<ACTION> <PROJECT>...Purpose Apply an action to one or more projects and start the Strategize phase.
Arguments
<ACTION_NAME>: Action name.<PROJECT>: One or more project names (positional arguments, repeatable).--arg/-a name=value: Action argument (repeatable).--automation-profile PROFILE: Automation profile name (e.g.,trusted,auto,local/careful-auto). Overrides the profile inherited from the action, project, or global config.--strategy-actor ACTOR: Override the action's strategy actor for this plan.--execution-actor ACTOR: Override the action's execution actor for this plan.--estimation-actor ACTOR: Override the action's estimation actor for this plan.--invariant-actor ACTOR: Override the action's Invariant Reconciliation Actor for this plan.--invariant TEXT: Invariant to attach to the created plan (repeatable). These are added as plan-level invariants in addition to any invariants inherited from the action, project, or global scope.
All actor arguments (--strategy-actor, --execution-actor, --estimation-actor, --invariant-actor) are optional overrides. When provided, they replace whatever was set when creating the action. When omitted, the action's configured actors are used.
Examples
$ agents plan use local/code-coverage local/api-service \ --arg target_coverage_percent=85 --automation-profile trusted╭─ Plan Created ──────────────────────╮ │ Plan ID: 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ │ Phase: strategize │ │ Action: local/code-coverage │ │ Project: local/api-service │ │ Automation: trusted │ │ Attempt: 1 │ ╰─────────────────────────────────────╯
╭─ Inputs ──────────────────────╮ │ - target_coverage_percent=85 │ │ - automation_profile=trusted │ ╰───────────────────────────────╯
╭─ Actors ────────────────────────╮ │ Strategy: local/strategist │ │ Execution: local/executor │ │ Estimation: (none) │ ╰─────────────────────────────────╯
╭─ Automation ─────────────────────────╮ │ Profile: trusted │ │ Source: CLI flag │ │ Read-Only: no │ ╰──────────────────────────────────────╯
╭─ Context ───────────────────────╮ │ Resources: 2 (repo, db) │ │ Indexed Files: 347 │ │ View: strategize │ │ Hot Token Budget: 12,000 │ ╰─────────────────────────────────╯
╭─ Next Steps ─────────────────────────────────────╮ │ - agents plan execute 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ │ - agents plan status 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ │ - agents plan tree 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ ╰──────────────────────────────────────────────────╯
✓ OK Plan created
Applying an action to multiple projects simultaneously, with plan-level invariants:
$ agents plan use local/security-audit local/api-service local/web-app \ --invariant "Never modify production database schemas" \ --invariant "All changes must include test coverage" \ --automation-profile supervised╭─ Plan Created ──────────────────────────────────────╮ │ Plan: 01HXM9D2ZK4Q7C2B3F2R4VYV6J │ │ Action: local/security-audit │ │ Phase: strategize (running) │ │ Automation: supervised │ ╰─────────────────────────────────────────────────────╯
╭─ Target Projects ───────────────────╮ │ 1. local/api-service (3 resources) │ │ 2. local/web-app (2 resources) │ ╰─────────────────────────────────────╯
╭─ Plan Invariants ──────────────────────────────────────────╮ │ Scope Source Invariant │ │ ──────── ─────── ────────────────────────────────── │ │ plan CLI Never modify production DB schemas │ │ plan CLI All changes must include test coverage │ │ project config API responses must be backward-compat │ │ global config Follow Python PEP 8 style guide │ ╰────────────────────────────────────────────────────────────╯
✓ OK Plan created — strategize in progress
Using an action with custom actor overrides:
$ agents plan use local/code-coverage local/api-service \ --strategy-actor local/senior-planner \ --execution-actor local/fast-executor \ --arg target_coverage_percent=95╭─ Plan Created ──────────────────────────────────────╮ │ Plan: 01HXM9E3ZK4Q7C2B3F2R4VYV6J │ │ Action: local/code-coverage │ │ Phase: strategize (running) │ │ Automation: trusted │ ╰─────────────────────────────────────────────────────╯
╭─ Actor Overrides ────────────────────╮ │ Strategy: local/senior-planner │ │ Execution: local/fast-executor │ │ (Estimation: action default) │ ╰──────────────────────────────────────╯
╭─ Arguments ─────────────────╮ │ target_coverage_percent: 95 │ ╰─────────────────────────────╯
✓ OK Plan created — strategize in progress
agents plan execute
agents plan execute <PLAN_ID>Purpose Start or resume execution for a plan.
Arguments
<PLAN_ID>: Plan ID (required).
Examples
$ agents plan execute 01HXM8C2ZK4Q7C2B3F2R4VYV6J╭─ Execution ──────────────────────╮ │ Plan: 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ │ Phase: execute │ │ Sandbox: git_worktree │ │ Worker: local/executor │ │ Started: 12:58:10 │ │ Attempt: 1 │ ╰──────────────────────────────────╯
╭─ Sandbox ──────────────────────────────────╮ │ Strategy: git_worktree │ │ Path: /repos/api/.worktrees/plan-01HXM8 │ │ Branch: cleveragents/plan-01HXM8C2 │ │ Status: active │ ╰────────────────────────────────────────────╯
╭─ Strategy Summary ─────────────────────╮ │ Decisions: 8 │ │ Invariants: 2 │ │ Planned Child Plans: 2+ │ │ Estimated Files: ~12 │ │ Risk: low │ ╰────────────────────────────────────────╯
╭─ Progress ────────╮ │ ⏳ Collect context │ │ • Run tools │ │ • Build changeset │ │ • Validate │ ╰───────────────────╯
✓ OK Execution started
Resuming execution of a plan that was previously paused or errored:
$ agents plan execute 01HXM7K2ZK4Q7C2B3F2R4VYV6J╭─ Execution Resumed ─────────────────╮ │ Plan: 01HXM7K2ZK4Q7C2B3F2R4VYV6J │ │ Phase: execute (resumed) │ │ Sandbox: git_worktree │ │ Worker: local/executor │ │ Checkpoint: cp_01HXM8C2 (loaded) │ │ Resumed From: step 4 of 6 │ ╰─────────────────────────────────────╯
╭─ Previous Progress ──────────────────────╮ │ ✓ Step 1: Collect context (0.8s) │ │ ✓ Step 2: Analyze codebase (4.2s) │ │ ✓ Step 3: Generate migrations (6.1s) │ │ ✗ Step 4: Apply migrations (errored) │ │ ○ Step 5: Update models (pending) │ │ ○ Step 6: Run validations (pending) │ ╰──────────────────────────────────────────╯
╭─ Guidance Applied ──────────────────────────────────────────────╮ │ "Use smaller batch sizes for the migration to avoid timeouts" │ ╰─────────────────────────────────────────────────────────────────╯
✓ OK Execution resumed from checkpoint cp_01HXM8C2
agents plan apply
agents plan apply [--yes|-y] <PLAN_ID>Purpose Apply sandboxed changes to real resources.
Arguments
<PLAN_ID>: Plan ID.--yes: Skip confirmation.
Examples
$ agents plan apply 01HXM8C2ZK4Q7C2B3F2R4VYV6JApply changes for plan 01HXM8C2ZK4Q7C2B3F2R4VYV6J? [y/N]: y
╭─ Apply Summary ─────────────────────╮ │ Plan: 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ │ Artifacts: 6 files updated │ │ Changes: 42 insertions, 9 deletions │ │ Project: local/api-service │ │ Applied At: 2026-02-08 13:04 │ ╰─────────────────────────────────────╯
╭─ Validation (from Execute) ────╮ │ Tests: passed (24/24) │ │ Lint: passed (0 warnings) │ │ Type Check: passed (0 errors) │ │ Duration: 12.4s │ ╰────────────────────────────────╯
╭─ Sandbox Cleanup ─────────╮ │ Worktree: removed │ │ Branch: merged to main │ │ Checkpoint: archived │ ╰───────────────────────────╯
╭─ Plan Lifecycle ────────────────────────╮ │ Phase: apply │ │ State: applied │ │ Total Duration: 00:06:14 │ │ Total Cost: $0.0847 │ │ Decisions Made: 8 │ │ Child Plans: 2 (completed) │ ╰─────────────────────────────────────────╯
╭─ Next Steps ──────╮ │ - Review git diff │ │ - Commit changes │ ╰───────────────────╯
✓ OK Changes applied
When Execute-phase validations failed and were not resolved (e.g., the execution actor exhausted its retry limit or failed outright), the plan cannot be applied. Apply checks the validation record and refuses to proceed:
$ agents plan apply --yes 01HXM8C2ZK4Q7C2B3F2R4VYV6J╭─ Apply Summary ─────────────────────╮ │ Plan: 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ │ Artifacts: 6 files updated │ │ Changes: 42 insertions, 9 deletions │ │ Project: local/api-service │ ╰─────────────────────────────────────╯
╭─ Validation ───────────────────────────────────────────────╮ │ ✗ Tests: FAILED (22/24 passed, 2 failed) │ │ FAIL test_auth.py::test_session_refresh — AssertionError │ │ FAIL test_auth.py::test_token_expiry — TimeoutError │ │ ✓ Lint: passed (0 warnings) │ │ ✓ Type Check: passed (0 errors) │ │ Duration: 14.8s │ ╰────────────────────────────────────────────────────────────╯
╭─ Sandbox Status ─────────────────────────────────────────────╮ │ Worktree: preserved (changes NOT committed) │ │ Checkpoint: cp_01HXM8C2 (pre-apply state available) │ │ The sandbox is preserved for correction or manual review. │ ╰──────────────────────────────────────────────────────────────╯
╭─ Recovery Options ──────────────────────────────────────────────────╮ │ - agents plan prompt — provide guidance to fix test failures │ │ - agents plan correct — revert and re-execute with guidance │ │ - agents plan rollback — restore to a previous checkpoint │ │ - agents plan cancel — abort the plan entirely │ ╰─────────────────────────────────────────────────────────────────────╯
✗ ERROR Apply refused — 2 required Execute-phase validations did not pass
agents plan status
agents plan status <PLAN_ID>Purpose Show detailed status for a plan.
Arguments
<PLAN_ID>: Plan ID (required).
Examples
$ agents plan status 01HXM8C2ZK4Q7C2B3F2R4VYV6J╭─ Plan Status ────────────────────╮ │ Plan: 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ │ Phase: execute │ │ State: processing │ │ Action: local/code-coverage │ │ Project: local/api-service │ │ Automation: review │ │ Attempt: 1 │ ╰──────────────────────────────────╯
╭─ Progress ───────╮ │ ✓ Strategize │ │ ⏳ Execute │ │ • Apply (queued) │ ╰──────────────────╯
╭─ Timing ──────────╮ │ Started: 12:57:01 │ │ Elapsed: 00:01:12 │ │ ETA: 00:03:45 │ ╰───────────────────╯
╭─ Execution Detail ──────────╮ │ Sandbox: git_worktree │ │ Tool Calls: 8 │ │ Files Modified: 3 │ │ Child Plans: 1/2 complete │ │ Checkpoints: 2 created │ ╰─────────────────────────────╯
╭─ Cost ───────────────╮ │ Tokens Used: 12,420 │ │ Cost So Far: $0.041 │ │ Estimated: $0.085 │ ╰──────────────────────╯
✓ OK Status refreshed
Status of a plan that has completed successfully:
$ agents plan status 01HXM6R3ZK4Q7C2B3F2R4VYV6J╭─ Plan Status ─────────────────────╮ │ Plan: 01HXM6R3ZK4Q7C2B3F2R4VYV6J │ │ Phase: apply │ │ State: applied │ │ Action: local/add-auth │ │ Project: local/api-service │ │ Automation: trusted │ │ Attempt: 1 │ ╰───────────────────────────────────╯
╭─ Progress ───────────╮ │ ✓ Strategize │ │ ✓ Execute │ │ ✓ Apply (committed) │ ╰──────────────────────╯
╭─ Timing ─────────────────────────╮ │ Started: 2026-02-08 12:57:01 │ │ Finished: 2026-02-08 13:04:15 │ │ Total Duration: 00:07:14 │ ╰──────────────────────────────────╯
╭─ Result ────────────────────────────╮ │ Decisions Made: 8 │ │ Child Plans: 2/2 complete │ │ Artifacts: 6 files updated │ │ Validations: 3/3 passed │ │ Total Cost: $0.085 │ ╰─────────────────────────────────────╯
✓ OK Plan completed successfully
Status of a plan in the strategize phase:
$ agents plan status 01HXM9F2ZK4Q7C2B3F2R4VYV6J╭─ Plan Status ────────────────────╮ │ Plan: 01HXM9F2ZK4Q7C2B3F2R4VYV6J │ │ Phase: strategize │ │ State: processing │ │ Action: local/refactor-auth │ │ Project: local/api-service │ │ Automation: supervised │ ╰──────────────────────────────────╯
╭─ Progress ───────────────╮ │ ⏳ Strategize (running) │ │ ○ Execute (waiting) │ │ ○ Apply (waiting) │ ╰──────────────────────────╯
╭─ Strategy Progress ────────╮ │ Decisions Made: 4 │ │ Invariants Enforced: 2 │ │ Child Plans Planned: 3 │ │ Elapsed: 00:00:28 │ ╰────────────────────────────╯
✓ OK Strategize in progress
Status of a plan that encountered an error:
$ agents plan status 01HXM7K2ZK4Q7C2B3F2R4VYV6J╭─ Plan Status ────────────────────╮ │ Plan: 01HXM7K2ZK4Q7C2B3F2R4VYV6J │ │ Phase: execute │ │ State: errored │ │ Action: local/migrate-db │ │ Project: local/api-service │ │ Automation: supervised │ │ Attempt: 1 │ ╰──────────────────────────────────╯
╭─ Progress ───────────╮ │ ✓ Strategize │ │ ✗ Execute │ │ ○ Apply (skipped) │ ╰──────────────────────╯
╭─ Error Detail ──────────────────────────────────────────────────╮ │ Error: Tool invocation failed: connection refused │ │ Tool: local/db-migrate │ │ Step: 4 of 6 │ │ Checkpoint: cp_01HXM8C2 (sandbox state preserved) │ │ Recoverable: yes — use "agents plan prompt" to provide guidance │ ╰─────────────────────────────────────────────────────────────────╯
╭─ Cost ───────────────╮ │ Tokens Used: 8,340 │ │ Cost So Far: $0.028 │ ╰──────────────────────╯
✗ ERROR Plan errored — useagents plan promptto resume oragents plan cancelto abort
agents plan cancel
agents plan cancel [(--reason|-r) <REASON>] <PLAN_ID>Purpose Cancel a plan that is not terminal.
Arguments
<PLAN_ID>: Plan ID.--reason/-r TEXT: Optional reason.
Examples
$ agents plan cancel 01HXM8C2ZK4Q7C2B3F2R4VYV6J --reason "blocked on credentials"╭─ Plan Cancelled ─────────────────╮ │ Plan: 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ │ Phase: execute │ │ Reason: blocked on credentials │ │ State: cancelled │ │ Cancelled At: 13:02:15 │ ╰──────────────────────────────────╯
╭─ Sandbox ────────────────╮ │ Status: preserved │ │ Files Modified: 3 │ │ Checkpoints: 2 │ ╰──────────────────────────╯
╭─ Child Plans ─────────────╮ │ Completed: 1 │ │ Cancelled: 1 │ │ Artifacts Preserved: yes │ ╰───────────────────────────╯
╭─ Recovery ───────────────────────────────────────────╮ │ - Resolve credentials │ │ - Run agents plan execute 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ ╰──────────────────────────────────────────────────────╯
✓ OK Plan cancelled
agents plan tree
agents plan tree [--show-superseded] <PLAN_ID>Purpose Render the decision tree for a plan.
Arguments
<PLAN_ID>: Plan ID (required).--show-superseded: Include superseded decisions.
Examples
$ agents plan tree 01HXM8C2ZK4Q7C2B3F2R4VYV6J╭─ Decision Tree ──────────────────────────────────────────────────────────────────────────╮ │ Plan: 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ │ ├─ [prompt_definition] "Increase test coverage to 85%" │ │ ├─ [invariant_enforced] "Prioritize financial transaction and user mgmt code" │ │ ├─ [invariant_enforced] "All API calls over TCP must be mocked" │ │ ├─ [strategy_choice] "Prioritize auth and payments" (confidence: 0.82) │ │ ├─ [subplan_parallel_spawn] "Implement auth and payment modules, in parallel" │ │ │ ├─ [subplan_spawn] "Write auth tests" → Plan: 01HXM9F1A │ │ │ └─ [subplan_spawn] "Write payment tests" → Plan: 01HXM9F2B │ │ └─ [subplan_parallel_spawn] "Write tests for remaining modules" │ │ └─ ... │ ╰──────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Tree Summary ─────────────╮ │ Nodes: 9 │ │ Depth: 3 │ │ Child Plans: 2+ │ │ Invariants: 2 │ │ Superseded: 0 (hidden) │ ╰────────────────────────────╯
╭─ Child Plans ──────────────────────────────────────╮ │ ID Phase State │ │ ────────── ─────── ───────── │ │ 01HXM9F1A execute processing │ │ 01HXM9F2B execute queued │ ╰────────────────────────────────────────────────────╯
╭─ Decision IDs (for correction) ──────────────────╮ │ Root: 01HXM9A0B1Q2W3R5G8Z0P4Q1X8 │ │ Invariant 1: 01HXM9A0C1R3X4S6G9Z1P5Q2Y9 │ │ Invariant 2: 01HXM9A0D2S4Y5T7H0Z2P6Q3Z0 │ │ Strategy: 01HXM9A1C2Q7W3R5G8Z0P4Q1X9 │ │ Parallel 1: 01HXM9A1D3R8X5S7H1Z2P5Q2Y0 │ │ Spawn Auth: 01HXM9A2D3Q8W4R6H9Z1P5Q2X0 │ │ Spawn Payment: 01HXM9A3E4Q9W5R7I0Z2P6Q3X1 │ │ Parallel 2: 01HXM9A4F5Q0W6R8J1Z3P7Q4X2 │ ╰──────────────────────────────────────────────────╯
✓ OK Decision tree rendered
agents plan explain
agents plan explain [--show-context] [--show-reasoning] <DECISION_ID>Purpose Show a detailed explanation for a decision.
Arguments
<DECISION_ID>: Decision ID.--show-context: Include the context snapshot.--show-reasoning: Include raw model reasoning if available.
Examples
$ agents plan explain 01HXM9A1C2Q7W3R5G8Z0P4Q1X9 --show-context╭─ Decision ─────────────────────────────────────╮ │ ID: 01HXM9A1C2Q7W3R5G8Z0P4Q1X9 │ │ Type: strategy_choice │ │ Question: Which modules should be prioritized? │ │ Chosen: Auth and payments │ │ Confidence: 0.82 │ │ Plan: 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ │ Sequence: 2 of 5 │ │ Created: 2026-02-08 12:58 │ ╰────────────────────────────────────────────────╯
╭─ Alternatives Considered ──────────────────────╮ │ 1. Auth and payments (chosen) │ │ 2. User module first (coverage 71%, med risk) │ │ 3. All modules equally (spread thin) │ ╰────────────────────────────────────────────────╯
╭─ Impact ──────────────────────╮ │ Downstream Decisions: 3 │ │ Downstream Child Plans: 2 │ │ Artifacts Produced: 5 │ │ Correction Impact: medium │ ╰───────────────────────────────╯
╭─ Context Snapshot ───────────────╮ │ - Coverage < 70% in auth │ │ - Payments failures last release │ │ - Auth: 12 files, 45% coverage │ │ - Payments: 8 files, 52% cover. │ │ Hot Context Hash: sha256:4b2e... │ ╰──────────────────────────────────╯
╭─ Rationale ───────────────────────────────────────╮ │ Auth and payment modules have the lowest coverage │ │ and highest business risk. Auth handles security │ │ tokens, payments handles money. Both had bugs in │ │ the last release traceable to missing tests. │ ╰───────────────────────────────────────────────────╯
╭─ Correction ──────────────────────────────────────────────╮ │ agents plan correct 01HXM9A1C2Q7W3R5G8Z0P4Q1X9 │ │ --mode revert --guidance "Prioritize payments first..." │ ╰───────────────────────────────────────────────────────────╯
✓ OK Decision explained
Including the raw model reasoning with --show-reasoning:
$ agents plan explain --show-reasoning 01HXM9A1C2Q7W3R5G8Z0P4Q1X9╭─ Decision ──────────────────────────────╮ │ ID: 01HXM9A1C2Q7W3R5G8Z0P4Q1X9 │ │ Type: strategy_choice │ │ Choice: Convert to async/await │ │ Alternatives: 3 │ ╰─────────────────────────────────────────╯
╭─ Alternatives Considered ────────────────────────────────────╮ │ 1. Convert to async/await patterns (chosen) │ │ 2. Keep synchronous with thread pool │ │ 3. Use callback-based approach │ ╰──────────────────────────────────────────────────────────────╯
╭─ Rationale ──────────────────────────────────────────────────╮ │ Async/await is the modern Python standard for I/O-bound │ │ operations. The project already uses asyncio in 3 modules. │ │ Thread pools would add complexity without native support. │ ╰──────────────────────────────────────────────────────────────╯
╭─ Model Reasoning (raw) ───────────────────────────────────────╮ │ I need to decide on the concurrency pattern for the payment │ │ processing module. Let me analyze the current codebase: │ │ │ │ 1. src/payments/api.py uses synchronous requests. │ │ 2. src/core/scheduler.py already uses asyncio. │ │ 3. The database driver (asyncpg) supports async natively. │ │ 4. Project invariant says "prefer modern Python patterns". │ │ │ │ Given that 3/5 core modules already use asyncio, and the │ │ database driver supports it, converting to async/await is │ │ the most consistent choice. Thread pools would work but │ │ add unnecessary complexity and don't integrate well with │ │ the existing asyncio event loop in scheduler.py. │ ╰───────────────────────────────────────────────────────────────╯
✓ OK Decision explained
agents plan correct
agents plan correct --mode (revert|append) (--guidance|-g) <GUIDANCE>
[--dry-run] [--yes|-y] <DECISION_ID>Purpose Correct a decision either by reverting and re-executing or by appending a fix.
Arguments
<DECISION_ID>: Decision ID.--mode revert|append: Correction mode.--guidance/-g TEXT: Guidance text.--dry-run: Show impact without executing.--yes: Skip confirmation for revert mode.
Examples
$ agents plan correct 01HXM9A1C2Q7W3R5G8Z0P4Q1X9 --mode revert \ --guidance "Prioritize payments first" --yes╭─ Correction ────────────────────────────────────────╮ │ Mode: revert │ │ Impact: 3 decisions, 2 child plans, 5 artifacts │ │ New Decision: 01HXM9B7Z3Q1Q8K2E9H7K3W2M8 │ │ Corrects: 01HXM9A1C2Q7W3R5G8Z0P4Q1X9 │ │ Attempt: 2 │ ╰─────────────────────────────────────────────────────╯
╭─ Affected Subtree ──────────────╮ │ Decisions Invalidated: 3 │ │ Child Plans Rolled Back: 2 │ │ Artifacts Archived: 5 │ │ Unaffected Decisions: 2 │ ╰─────────────────────────────────╯
╭─ Sandbox Rollback ─────────────╮ │ Checkpoint: cp_01HXM8C2 │ │ Files Reverted: 5 │ │ Status: restored │ ╰────────────────────────────────╯
╭─ Recompute ──────────────╮ │ Queued: 2 child plans │ │ ETA: 4m │ ╰──────────────────────────╯
╭─ History ───────────────────────────────────────────╮ │ - Original decision superseded │ │ - Prior artifacts archived for comparison │ │ - agents plan diff --correction 01HXM9B7Z3Q1Q8K2.. │ ╰─────────────────────────────────────────────────────╯
✓ OK Correction applied
Using --mode append to add a corrective decision without reverting existing work (useful when the original decision was partially correct):
$ agents plan correct 01HXM9A1C2Q7W3R5G8Z0P4Q1X9 --mode append \ --guidance "Also add rate limiting to the auth endpoints"╭─ Correction ─────────────────────────────────────╮ │ Mode: append │ │ Impact: adds to existing subtree, no rollback │ │ New Decision: 01HXM9C3Z5T2Q8K2E9H7K3W2M8 │ │ Appended After: 01HXM9A1C2Q7W3R5G8Z0P4Q1X9 │ │ Attempt: 2 │ ╰──────────────────────────────────────────────────╯
╭─ Append Detail ─────────────────────────────────────────────────╮ │ Original decision preserved: yes │ │ Existing artifacts kept: yes │ │ Additional work: appended as new child plan │ │ The original 5 artifacts remain; a new child plan will add │ │ rate-limiting code on top of the existing auth changes. │ ╰─────────────────────────────────────────────────────────────────╯
╭─ Queued ──────────╮ │ New child: 1 │ │ ETA: 2m │ ╰───────────────────╯
✓ OK Append correction queued
Using --dry-run to preview the impact of a correction before committing to it:
$ agents plan correct 01HXM9A1C2Q7W3R5G8Z0P4Q1X9 --mode revert \ --guidance "Use async/await pattern instead" --dry-run╭─ Dry Run — Correction Preview ─────────────────────────────────────╮ │ ⚠ This is a preview only. No changes will be made. │ ╰────────────────────────────────────────────────────────────────────╯
╭─ Would Revert ───────────────────────────────────────╮ │ Decisions to invalidate: 3 │ │ 01HXM9A1.. strategy_choice "sync pattern" │ │ 01HXM9A2.. implementation_choice "requests" │ │ 01HXM9A3.. tool_invocation write_file ×4 │ │ Child plans to roll back: 2 │ │ Artifacts to archive: 5 files │ │ Unaffected decisions: 2 (will be kept) │ ╰──────────────────────────────────────────────────────╯
╭─ Estimated Cost ──────────╮ │ Re-strategize: ~$0.012 │ │ Re-execute: ~$0.035 │ │ Total: ~$0.047 │ │ ETA: ~4 minutes │ ╰───────────────────────────╯
To execute this correction, remove --dry-run and add --yes
agents plan diff
agents plan diff (--correction <CORRECTION_ATTEMPT_ID>|<PLAN_ID>)Purpose Show diffs for a plan or a correction attempt.
Arguments
<PLAN_ID>: Show diff for a plan (positional argument). Mutually exclusive with --correction.--correction CORRECTION_ATTEMPT_ID: Show diff for a correction attempt instead.
Examples
$ agents plan diff 01HXM8C2ZK4Q7C2B3F2R4VYV6J╭─ Diff Summary ─────────────────────────────────╮ │ Plan: 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ │ Project: local/api-service │ │ Files Changed: 2 │ │ Insertions: 12 │ │ Deletions: 4 │ │ Net Change: +8 lines │ ╰────────────────────────────────────────────────╯
╭─ Files ───────────────────────────────╮ │ Path Change Status │ │ ─────────────────── ────── ──────── │ │ src/auth/session.py +8 -2 modified │ │ src/auth/tokens.py +4 -2 modified │ ╰───────────────────────────────────────╯
╭─ Patch Preview ──────────────────────────╮ │ --- a/src/auth/session.py │ │ +++ b/src/auth/session.py │ │ @@ -12,4 +12,10 @@ │ │ - import jwt │ │ + import sessionlib │ │ - def validate_token(...) │ │ + def validate_session(...) │ │ --- a/src/auth/tokens.py │ │ +++ b/src/auth/tokens.py │ │ @@ -5,3 +5,7 @@ │ │ - TOKEN_EXPIRY = 3600 │ │ + TOKEN_EXPIRY = 7200 │ ╰──────────────────────────────────────────╯
╭─ Risk Assessment ────────────────╮ │ API Compatibility: preserved │ │ Test Coverage: maintained │ │ Breaking Changes: none detected │ ╰──────────────────────────────────╯
✓ OK Diff generated
Showing the diff for a specific correction attempt, comparing what changed between the original and corrected execution:
$ agents plan diff --correction 01HXM9B7Z3Q1Q8K2E9H7K3W2M8╭─ Correction Diff ───────────────────────────────╮ │ Correction: 01HXM9B7Z3Q1Q8K2E9H7K3W2M8 │ │ Original Decision: 01HXM9A1C2Q7W3R5.. │ │ Mode: revert │ │ Files Changed: 3 │ │ New Insertions: 18 │ │ New Deletions: 6 │ ╰─────────────────────────────────────────────────╯
╭─ Comparison ─────────────────────────────────────────────────────╮ │ File Before (original) After (corrected) │ │ ──────────────────── ──────────────── ──────────────────── │ │ src/payments/api.py +12 -4 +18 -6 (expanded) │ │ src/auth/tokens.py +8 -2 (unchanged) │ │ tests/test_payments.py (new file) (new file, larger) │ ╰──────────────────────────────────────────────────────────────────╯
╭─ Patch Preview (corrected vs original) ───────────╮ │ --- a/src/payments/api.py (original) │ │ +++ b/src/payments/api.py (corrected) │ │ @@ -1,12 +1,18 @@ │ │ - # sync payment processing │ │ + # async payment processing (corrected) │ │ + import asyncio │ │ + from aiohttp import ClientSession │ │ ... │ ╰───────────────────────────────────────────────────╯
✓ OK Correction diff generated
agents plan artifacts
agents plan artifacts <PLAN_ID>Purpose List artifacts produced by a plan.
Arguments
<PLAN_ID>: Plan ID.
Examples
$ agents plan artifacts 01HXM8C2ZK4Q7C2B3F2R4VYV6J╭─ Artifacts ─────────────────────────────────────────────────╮ │ Path Type Size Change Child Plan │ │ ───────────────────── ───── ────── ───────── ─────── │ │ src/auth/session.py write 2.1 KB +8 -2 root │ │ tests/test_session.py write 4.7 KB +47 -0 root │ │ src/auth/tokens.py edit 1.8 KB +4 -2 root │ │ tests/test_tokens.py write 3.2 KB +32 -0 auth │ ╰─────────────────────────────────────────────────────────────╯
╭─ Summary ───────────╮ │ Total: 4 │ │ Writes: 2 (new) │ │ Edits: 2 (modified) │ │ Deletes: 0 │ │ Total Size: 11.8 KB │ ╰─────────────────────╯
╭─ By Plan ─────────────────╮ │ Root Plan: 3 artifacts │ │ auth-tests: 1 artifact │ │ payment-tests: (pending) │ ╰───────────────────────────╯
✓ OK 4 artifacts listed
agents plan prompt
agents plan prompt <PLAN_ID> <GUIDANCE>Purpose Provide additional guidance to a plan, typically when it is errored or awaiting input.
Arguments
<PLAN_ID>: Plan ID."<GUIDANCE>": Guidance text.
Examples
$ agents plan prompt 01HXM8C2ZK4Q7C2B3F2R4VYV6J "Use mocks for database tests"╭─ Guidance Added ────────────────────────╮ │ Plan: 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ │ Guidance: Use mocks for database tests │ │ Scope: next execution step │ │ Phase: execute │ │ State: errored → processing │ ╰─────────────────────────────────────────╯
╭─ Decision Created ────────────────────────╮ │ Type: user_intervention │ │ ID: 01HXM9C5G7R2X8S3K4Z5Q8R6Y3 │ │ Parent: 01HXM9A1C2Q7W3R5G8Z0P4Q1X9 │ ╰───────────────────────────────────────────╯
╭─ Queue ────╮ │ Pending: 1 │ │ Applied: 0 │ ╰────────────╯
✓ OK Guidance queued
agents plan rollback
agents plan rollback [--yes|-y] <PLAN_ID> <CHECKPOINT_ID>Purpose Rollback a plan sandbox to a checkpoint.
Arguments
<PLAN_ID>: Plan ID.<CHECKPOINT_ID>: Checkpoint ID.--yes: Skip confirmation.
Examples
$ agents plan rollback 01HXM8C2ZK4Q7C2B3F2R4VYV6J cp_01HXM8C2Rollback plan 01HXM8C2ZK4Q7C2B3F2R4VYV6J to cp_01HXM8C2? [y/N]: y
╭─ Rollback Summary ───────────────╮ │ Plan: 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ │ Checkpoint: cp_01HXM8C2 │ │ Label: before auth refactor │ │ Files: 6 reverted │ ╰──────────────────────────────────╯
╭─ Changes Reverted ──────────────────╮ │ File Action │ │ ────────────────────── ────────── │ │ src/auth/session.py restored │ │ src/auth/tokens.py restored │ │ tests/test_session.py removed │ │ tests/test_tokens.py removed │ │ src/auth/fixtures.py restored │ │ src/auth/__init__.py restored │ ╰─────────────────────────────────────╯
╭─ Impact ──────────────────────────────╮ │ Child Plans Invalidated: 2 │ │ Sandbox: restored to cp_01HXM8C2 │ │ Decisions After CP: 2 discarded │ │ Tool Calls After CP: 5 undone │ ╰───────────────────────────────────────╯
╭─ Post-Rollback State ──────────╮ │ Phase: execute │ │ State: queued (awaiting input) │ │ Checkpoints Remaining: 2 │ ╰────────────────────────────────╯
✓ OK Rollback complete
agents action
Purpose Manage reusable actions.
agents action create
agents action create (--config|-c) <CFG_FILE>Purpose
Create a new action template from a YAML configuration file. The --config file is required and must fully define the action.
Arguments
--config/-c FILE: YAML configuration file defining the action (required). The file must fully define the action, including thenamefield,strategy-actor,execution-actor, anddefinition-of-done.
Examples
$ agents action create --config ./actions/code-coverage.yaml╭─ Action Created ────────────────────────╮ │ Name: local/code-coverage │ │ State: available │ │ Strategy Actor: local/strategist │ │ Execution Actor: local/executor │ │ Reusable: yes │ │ Read Only: no │ │ Config: ./actions/code-coverage.yaml │ │ Created: 2026-02-08 12:20 │ ╰─────────────────────────────────────────╯
╭─ Definition of Done ─╮ │ Coverage reaches 85% │ ╰──────────────────────╯
╭─ Arguments ──────────────────────────────────────────────────────╮ │ Name Type Required Description │ │ ─────────────────────── ────── ──────── ───────────────────── │ │ target_coverage_percent int yes Target coverage % │ │ test_command string no Test framework to use │ ╰──────────────────────────────────────────────────────────────────╯
╭─ Automation ──────────────────────────╮ │ Profile: supervised │ │ Source: default │ ╰───────────────────────────────────────╯
✓ OK Action created
agents action list
agents action list [(--namespace|-n) <NS>] [(--state|-s) <STATE>] [<REGEX>]Purpose List actions with optional filters.
Arguments
--namespace/-n NS: Filter by namespace.--state/-s STATE: Filter by state.
Examples
$ agents action list╭─ Actions ──────────────────────────────────────────────────────────────────────────────────╮ │ Name State Strategy Actor Execution Actor Reusable Plans │ │ ─────────────────── ───────── ──────────────── ─────────────── ──────── ───── │ │ local/code-coverage available local/strategist local/executor ✓ 3 │ ╰────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Filters ────────╮ │ State: available │ │ Namespace: (any) │ ╰──────────────────╯
╭─ Summary ──────────────╮ │ Total: 1 │ │ Available: 1 │ │ Archived: 0 │ │ Total Plans Created: 3 │ ╰────────────────────────╯
✓ OK 1 action listed
agents action show
agents action show <ACTION_NAME>Purpose Show details for an action.
Arguments
<ACTION_NAME>: Action name.
Examples
$ agents action show local/code-coverage╭─ Action Details ──────────────────────╮ │ Name: local/code-coverage │ │ State: available │ │ Strategy Actor: local/strategist │ │ Execution Actor: local/executor │ │ Reusable: yes │ │ Read Only: no │ │ Created: 2026-02-08 12:20 │ ╰───────────────────────────────────────╯
╭─ Definition of Done ─╮ │ Coverage reaches 85% │ ╰──────────────────────╯
╭─ Arguments ──────────────────────────────────────────────────────╮ │ Name Type Required Description │ │ ─────────────────────── ────── ──────── ───────────────────── │ │ target_coverage_percent int yes Target coverage % │ │ test_command string no Test framework to use │ ╰──────────────────────────────────────────────────────────────────╯
╭─ Automation ──────────────────────────╮ │ Profile: supervised │ │ Source: default │ ╰───────────────────────────────────────╯
╭─ History ────────────────────╮ │ Plans Created: 3 │ │ Plans Completed: 2 │ │ Plans Failed: 0 │ │ Avg Duration: 00:04:30 │ │ Avg Cost: $0.072 │ ╰──────────────────────────────╯
╭─ Usage ───────────────────────────────────────────────────────────╮ │ - agents plan use local/code-coverage local/api-service │ │ --arg target_coverage_percent=85 │ ╰───────────────────────────────────────────────────────────────────╯
✓ OK Action loaded
agents action archive
agents action archive <ACTION_NAME>Purpose Archive an action.
Arguments
<ACTION_NAME>: Action name.
Examples
$ agents action archive local/old-action╭─ Action Archived ───────────╮ │ Name: local/old-action │ │ State: available → archived │ │ Archived: 2026-02-08 12:22 │ ╰─────────────────────────────╯
╭─ Impact ───────────────────────╮ │ Availability: hidden from list │ │ Existing Plans: unchanged │ │ Active Plans: 0 affected │ ╰────────────────────────────────╯
╭─ History ─────────────────╮ │ Total Plans: 5 │ │ Completed: 4 │ │ Failed: 1 │ │ Last Used: 2026-02-06 │ ╰───────────────────────────╯
✓ OK Action archived
agents automation-profile
Purpose
Manage automation profiles — named collections of confidence thresholds (floating-point values from 0.0 to 1.0) that control which tasks are automated vs. require human approval. Each threshold specifies the minimum confidence level at which the system proceeds automatically; below the threshold, the system drops to manual mode. A threshold of 0.0 means "always automatic" and 1.0 means "always manual." Built-in profiles (manual, review, supervised, cautious, trusted, auto, ci, full-auto) are always available. Custom profiles follow the same <namespace>/<name> naming convention as other entities.
agents automation-profile add
agents automation-profile add (--config|-c) <FILE> [--update]Purpose
Register a new custom automation profile from a YAML configuration file. The profile is fully defined by the config file. If a profile with the same name already exists, the command fails unless the --update flag is provided.
Arguments
--config/-c FILE: YAML configuration file defining the profile (required). The file must fully define the profile, including thenamefield which determines the profile's registered name.--update: Replace an existing profile with the same name.
Examples
$ agents automation-profile add --config ./profiles/careful-auto.yaml╭─ Profile Registered ─────────────────────────────────────────────╮ │ Name: local/careful-auto │ │ Description: Autonomous execution with mandatory sandbox │ │ and manual apply │ │ Created: 2026-02-08 14:30 │ ╰──────────────────────────────────────────────────────────────────╯
╭─ Confidence Thresholds ────────────────╮ │ auto_strategize: 0.0 │ │ auto_execute: 0.0 │ │ auto_apply: 1.0 │ │ auto_decisions_strategize: 0.0 │ │ auto_decisions_execute: 0.0 │ │ auto_validation_fix: 0.0 │ │ auto_strategy_revision: 1.0 │ │ auto_child_plans: 0.0 │ │ auto_retry_transient: 0.0 │ │ auto_checkpoint_restore: 0.0 │ │ require_sandbox: true │ │ require_checkpoints: true │ │ allow_unsafe_tools: false │ ╰────────────────────────────────────────╯
✓ OK Profile registered
agents automation-profile remove
agents automation-profile remove [--yes|-y] <NAME>Purpose Remove a custom automation profile. Built-in profiles cannot be removed.
Arguments
<NAME>: Profile name.--yes, -y: Skip confirmation prompt.
Examples
$ agents automation-profile remove local/careful-autoRemove automation profile local/careful-auto? [y/N]: y
╭─ Profile Removed ──────────╮ │ Name: local/careful-auto │ ╰────────────────────────────╯
✓ OK Profile removed
agents automation-profile list
agents automation-profile list [<REGEX>]Purpose List all available automation profiles (built-in and custom).
Arguments
[REGEX]: Optional filter pattern.
Examples
$ agents automation-profile list╭─ Automation Profiles ──────────────────────────────────────────────────────────╮ │ Name Source Auto-Apply Sandbox Description │ │ ────────────────── ──────── ────────── ─────── ──────────────────────── │ │ manual built-in 1.0 yes Human-driven (default) │ │ review built-in 1.0 yes Auto-phase, manual decisions │ │ supervised built-in 1.0 yes Auto-plan, manual exec │ │ cautious built-in 1.0 yes Confidence-gated automation │ │ trusted built-in 1.0 yes Auto-exec, manual apply │ │ auto built-in 1.0 yes Full auto except apply │ │ ci built-in 0.0 yes Full auto, safe CI/CD │ │ full-auto built-in 0.0 no Complete automation │ │ local/careful-auto custom 0.8 yes Custom careful profile │ ╰────────────────────────────────────────────────────────────────────────────────╯
╭─ Summary ───────────╮ │ Built-in: 8 │ │ Custom: 1 │ │ Total: 9 │ ╰─────────────────────╯
✓ OK 9 profiles listed
agents automation-profile show
agents automation-profile show <NAME>Purpose Show full details for an automation profile, including all flag values.
Arguments
<NAME>: Profile name.
Examples
$ agents automation-profile show trusted╭─ Automation Profile ───────────────────────────────────────────╮ │ Name: trusted │ │ Source: built-in │ │ Description: Auto-exec, manual apply. Day-to-day development │ ╰────────────────────────────────────────────────────────────────╯
╭─ Phase Transitions (thresholds) ─────╮ │ auto_strategize: 0.0 │ │ auto_execute: 0.0 │ │ auto_apply: 1.0 │ ╰──────────────────────────────────────╯
╭─ Decision Automation (thresholds) ───╮ │ auto_decisions_strategize: 0.0 │ │ auto_decisions_execute: 0.0 │ ╰──────────────────────────────────────╯
╭─ Self-Repair (thresholds) ───────────╮ │ auto_validation_fix: 0.0 │ │ auto_strategy_revision: 1.0 │ │ auto_retry_transient: 0.0 │ │ auto_checkpoint_restore: 1.0 │ ╰──────────────────────────────────────╯
╭─ Execution Controls (thresholds) ────╮ │ auto_child_plans: 0.0 │ │ require_sandbox: true │ │ require_checkpoints: true │ │ allow_unsafe_tools: false │ ╰──────────────────────────────────────╯
✓ OK Profile loaded
agents config
Purpose Manage global configuration values.
agents config set
agents config set <key> <value>Purpose Set a configuration key.
Arguments
<key>: Hierarchical dot-path key name (e.g.,core.automation-profile,core.log.level,actor.default.invariant,core.format). See Global Configuration Keys for the complete reference.<value>: Value to set.
The actor.default.invariant key sets the default Invariant Reconciliation Actor used globally. This actor is used to reconcile invariant conflicts when neither the plan nor the project specifies one.
The core.format key sets the default output rendering format used by all commands. Accepted values are rich, color, table, plain, json, yaml. When set, this value is used unless overridden by the --format CLI flag. See Output Rendering Framework for details.
Examples
$ agents config set core.automation-profile trusted╭─ Config Updated ──────────────────────╮ │ Key: core.automation-profile │ │ Value: trusted │ │ Previous: manual │ │ Source: config │ │ Scope: global │ ╰───────────────────────────────────────╯
╭─ Effective ────────────────────────╮ │ Sessions: new sessions │ │ Plans: future plans (unless set) │ │ Existing: unchanged │ ╰────────────────────────────────────╯
╭─ Saved To ──────────────────────────╮ │ File: ~/.cleveragents/config.toml │ │ Line: 8 │ ╰─────────────────────────────────────╯
✓ OK Config updated
Setting the default output format:
$ agents config set core.format table╭─ Config Updated ─────────────────╮ │ Key: core.format │ │ Previous: rich │ │ New Value: table │ │ Scope: user (~/.cleveragents) │ ╰──────────────────────────────────╯
✓ OK Set core.format = table
Setting a global invariant actor:
$ agents config set actor.default.invariant local/invariant-resolver╭─ Config Updated ──────────────────────────────────╮ │ Key: actor.default.invariant │ │ Previous: (not set) │ │ New Value: local/invariant-resolver │ │ Scope: user (~/.cleveragents) │ ╰───────────────────────────────────────────────────╯
✓ OK Set actor.default.invariant = local/invariant-resolver
agents config get
agents config get <key>Purpose Get a configuration value.
Arguments
<key>: Key to read.
Examples
$ agents config get core.automation-profile╭─ Config ──────────────────────────╮ │ Key: core.automation-profile │ │ Value: trusted │ │ Source: config │ │ Overridden: no │ │ Type: string │ ╰───────────────────────────────────╯
╭─ Origin ──────────────────────────╮ │ File: ~/.cleveragents/config.toml │ │ Line: 8 │ │ Default: supervised │ ╰───────────────────────────────────╯
╭─ Resolution Chain ──────────────╮ │ 1. CLI flag: (not set) │ │ 2. Env var: (not set) │ │ 3. Config file: trusted │ │ 4. Default: supervised │ │ Winner: config file (level 3) │ ╰─────────────────────────────────╯
✓ OK Config read
agents config list
agents config list [--filter-values <REGEX>] [<REGEX>]Purpose List all configuration values.
Arguments
None.
Examples
$ agents config list╭─ Config ───────────────────────────────────────────────────╮ │ Key Value Source Modified │ │ ──────────────── ────────────────── ─────── ──────── │ │ core.automation-profile trusted config yes │ │ actor.default.invariant local/reconciler config yes │ │ core.log.level FATAL default no │ ╰────────────────────────────────────────────────────────────╯
╭─ Overrides ─────╮ │ Env: none │ │ CLI Flags: none │ ╰─────────────────╯
╭─ Config File ───────────────────────╮ │ Path: ~/.cleveragents/config.toml │ │ Size: 284 bytes │ │ Valid: yes │ ╰─────────────────────────────────────╯
✓ OK 6 settings listed
agents invariant
Purpose
Manage invariants — named constraints that guide and constrain plan execution. Invariants can be attached to any scope: global (all plans), project (all plans targeting that project), plan (a specific plan and its child plans), or action (carried forward when the action is used). Exactly one scope flag is required for add and list.
agents invariant add
agents invariant add [--global] [(--project|-p) PROJECT] [--plan PLAN_ID]...
[--action ACTION]... <INVARIANT_TEXT>Purpose Add an invariant at the specified scope.
Arguments
<INVARIANT_TEXT>: The invariant text (positional argument at end of command).--global: Attach as a global invariant (applies to all plans).--project/-p PROJECT: Attach to a project (applies to all plans targeting this project).--plan PLAN_ID: Attach to a plan (plan-level invariant). Repeatable.--action ACTION: Attach to an action (action-level invariant). Repeatable.
At least one scope flag (--global, --project, --plan, or --action) must be provided. --plan and --action can be repeated to attach the same invariant to multiple plans or actions.
Examples
$ agents invariant add --global "All public APIs must maintain backward compatibility"╭─ Invariant Added ──────────────────────────────────────────────────────╮ │ Invariant: All public APIs must maintain backward compatibility │ │ Scope: global │ │ ID: inv_01HXM9A1B │ ╰────────────────────────────────────────────────────────────────────────╯
✓ OK Invariant added
$ agents invariant add --project local/api-service "All endpoints must validate auth tokens"
╭─ Invariant Added ──────────────────────────────────────────────╮ │ Project: local/api-service │ │ Invariant: All endpoints must validate auth tokens │ │ Scope: project │ │ ID: inv_01HXM9A2C │ ╰────────────────────────────────────────────────────────────────╯
✓ OK Invariant added
$ agents invariant add --plan 01HXM8C2ZK4Q7C2B3F2R4VYV6J "All database queries must use parameterized statements"
╭─ Invariant Added ─────────────────────────────────────────────────────╮ │ Plan: 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ │ Invariant: All database queries must use parameterized statements │ │ Scope: plan │ │ ID: inv_01HXM9G3A │ ╰───────────────────────────────────────────────────────────────────────╯
✓ OK Invariant added
$ agents invariant add --action local/code-coverage "Test files must not import production secrets"
╭─ Invariant Added ──────────────────────────────────────────────────────╮ │ Action: local/code-coverage │ │ Invariant: Test files must not import production secrets │ │ Scope: action │ │ ID: inv_01HXM9H4B │ ╰────────────────────────────────────────────────────────────────────────╯
✓ OK Invariant added
agents invariant list
agents invariant list [--global] [(--project|-p) PROJECT] [--plan PLAN_ID] [--action ACTION]
[--effective] [<REGEX>]Purpose
List invariants at a given scope. Use --effective with --plan to show the final reconciled view of invariants (after precedence resolution) rather than just the invariants directly attached at that scope.
Arguments
--global: List global invariants.--project/-p PROJECT: List invariants attached to a project.--plan PLAN_ID: List invariants attached to a plan.--action ACTION: List invariants attached to an action.--effective: (Only with--plan) Show the reconciled invariant view after precedence resolution across all scopes.
Examples
$ agents invariant list --global╭─ Global Invariants ──────────────────────────────────────────────────────╮ │ ID Text │ │ ────────────── ──────────────────────────────────────────────────── │ │ inv_01HXM9A1B All public APIs must maintain backward compatibility │ │ inv_01HXM9A1C Payment processing must be idempotent │ ╰──────────────────────────────────────────────────────────────────────────╯
✓ OK 2 invariants
$ agents invariant list --plan 01HXM8C2ZK4Q7C2B3F2R4VYV6J --effective
╭─ Effective Invariants (Plan 01HXM8C2ZK4Q7C2B3F2R4VYV6J) ──────────────────────────────────╮ │ ID Source Text │ │ ────────────── ─────── ────────────────────────────────────────────────────── │ │ inv_01HXM9A1B global All public APIs must maintain backward compatibility │ │ inv_01HXM9A2C project All endpoints must validate auth tokens │ │ inv_01HXM9G3A plan All database queries must use parameterized statements │ ╰───────────────────────────────────────────────────────────────────────────────────────────╯
│ Conflicts Resolved: 1 │ │ Global "Use shared DB pool" overridden by plan "All database queries must use │ │ parameterized statements" │
✓ OK 3 effective invariants (1 global, 1 project, 1 plan; 1 conflict resolved)
agents invariant remove
agents invariant remove [--yes|-y] <INVARIANT_ID>Purpose Remove an invariant by ID. The invariant is removed from whichever scope it was originally attached to.
Arguments
<INVARIANT_ID>: Invariant ID to remove.--yes: Skip confirmation.
Examples
$ agents invariant remove inv_01HXM9A1CRemove invariant inv_01HXM9A1C ("Payment processing must be idempotent", scope: global)? [y/N]: y
╭─ Invariant Removed ──────────────────────────────────────────────╮ │ Removed: Payment processing must be idempotent │ │ Scope: global │ │ ID: inv_01HXM9A1C │ ╰──────────────────────────────────────────────────────────────────╯
✓ OK Invariant removed
Core Concepts
Plan
!!! adr "Architecture Decision" The plan lifecycle, phase transitions, and plan hierarchy are defined in ADR-006: Plan Lifecycle.
A plan is the fundamental unit of orchestration and traceability.
Plan Lifecycle Phases
A plan progresses through four phases:
Action → Strategize → Execute → Apply
In this spec:
- Action is the first phase. An action serves as a reusable plan template — it defines the work to be done (description, definition of done, actors, arguments, invariants) without being bound to any project. No significant processing occurs during the Action phase. When an action is used (via
agents plan use), it spawns a new plan entity that enters the Strategize phase. By default, the original action remains available for reuse; single-use actions (created withreusable: false) are deleted after first use. Although Action is formally a phase in the lifecycle, it is unique in that it functions primarily as a template from which plans are instantiated. - Strategize is the second phase and the first where active processing occurs (the output is a strategy).
- Execute is the third phase (the output is a changeset).
- Apply is the fourth and final phase. Apply merges the sandbox changeset into real project resources. Upon completion, the plan reaches one of several terminal states:
applied(success — changes committed),constrained(cannot complete within the current strategy's constraints — may trigger reversion to Strategize),errored(failed), orcancelled.
The normal phase progression is forward, but both Execute and Apply may revert to Strategize when the current strategy's constraints are too restrictive (see Phase Reversion below).
Phase Transition Verbs (CLI / UX Contract)
Verbs that trigger phase transitions. CleverAgents should standardize these verbs as the public API (CLI, TUI, web):
| Current Phase | Command Verb | Next Phase |
|---|---|---|
| (none) | create |
Action |
| Action | use |
Strategize |
| Strategize | execute |
Execute |
| Execute | apply |
Apply |
Important behavioral rules:
- CleverAgents uses automation profiles to control which of these phase transitions happen automatically. The profile determines whether each transition requires explicit user action or proceeds autonomously, but the verbs remain the conceptual contract.
- There is no
strategizecommand. The phase transition verbs are used in CLI commands — not the phase names. Theuseverb (as inagents plan use) is the command that transitions a plan from the Action phase into Strategize. This is by design: "use" describes the user's intent (using an action template on a project), and the system responds by entering the Strategize phase.
Phase Reversion (→ Strategize)
!!! adr "Architecture Decision" Phase reversion rules and the decision tree model are detailed in ADR-006: Plan Lifecycle and ADR-007: Decision Tree and Correction.
While the normal phase progression is forward (Action → Strategize → Execute → Apply), both the Execute and Apply phases may revert to Strategize under specific conditions. In all cases, the reversion target is always Strategize — there is no reversion to Execute or any other intermediate phase.
Execute → Strategize
- Trigger: During execution, the execution actor discovers that constraints established during Strategize are too restrictive to complete the work. For example, a strategy decision may have mandated an approach that proves infeasible once actual resources are examined, dependencies are resolved, or implementation details are encountered.
- Behavior: Rather than overriding or silently violating Strategize-phase decisions, the plan reverts to the Strategize phase so the strategy actor can adjust the decision tree. Execute-phase decisions must always be constrained by Strategize-phase decisions — if those constraints cannot be honored, the correct response is reversion, not violation.
- After reversion: The strategy actor receives the execution actor's findings (the reason for reversion, the specific constraints that were too restrictive) and produces an updated decision tree. The plan then re-enters the Execute phase with the revised strategy.
- Automation: Controlled by the
auto_strategy_revisionautomation profile flag.
Apply → Strategize
- Trigger: During the Apply phase, the system determines that the changeset cannot be successfully applied within the current strategy's constraints. Examples include: merge conflicts that violate invariants, validation failures at apply time that cannot be resolved within the strategy's scope, or resource state drift that contradicts strategy assumptions.
- Behavior: The plan enters the
constrainedterminal state within the Apply phase. From this state, the plan may revert to Strategize either automatically (if theauto_reversion_from_applyautomation profile flag allows it) or after manual user approval. - After reversion: The strategy actor receives the Apply phase's findings (the constraint violations, the specific issues encountered) and produces an updated decision tree. The plan then progresses forward again through Execute and Apply with the revised strategy.
- Automation: Controlled by the
auto_reversion_from_applyautomation profile flag (separate fromauto_strategy_revision, which governs Execute → Strategize reversion).
Plan States (Per Phase)
A plan's phase indicates "what step of the lifecycle it is in." Separately, the plan has a processing state indicating "what is happening right now."
Recommended state model:
-
Action phase states
available(action exists and can be used)archived(soft-deleted or hidden, optional)
-
Strategize / Execute phase states
queued(waiting for compute/worker)processing(currently running)errored(failed; includes error metadata)complete(finished successfully)cancelled(user/system cancelled; safe terminal for that phase)
-
Apply phase states
queued(waiting for compute/worker)processing(currently running — diff review, conflict resolution, validation)errored(failed; includes error metadata)applied(terminal — changes successfully committed to real resources)constrained(terminal — cannot complete within the current strategy's constraints; may trigger automatic or manual reversion to Strategize, controlled by theauto_reversion_from_applyautomation profile flag)cancelled(user/system cancelled; safe terminal)
Plan Identity and Traceability
Every plan should have:
- plan_id: Unique, immutable ID (UUID or ULID).
- parent_plan_id: Nullable; present for child plans.
- root_plan_id: The top-most plan in the tree.
- attempt: An integer attempt counter that increments when re-running a phase (e.g., re-executing after a fix).
- created_at / updated_at / completed_at timestamps.
- created_by (user identity / session identity).
Plan Hierarchy and Parallelism
!!! adr "Architecture Decision" Plan hierarchy, child plan spawning, and parallel execution are covered in ADR-006: Plan Lifecycle.
A single plan should usually represent the smallest "complete" unit of work (similar to what would fit in one git commit). However:
- Plans are hierarchical. A child plan is simply a Plan with a parent — it follows the same lifecycle, has the same data model, and is functionally identical to a root plan except that it has a
parent_plan_id. - Decisions about child plans are made during Strategize (as
subplan_spawndecision types). Multiple child plans that should execute concurrently are grouped under asubplan_parallel_spawndecision. - Child plans are actually spawned during Execute (based on those decisions).
- Child plans can run in parallel (via
subplan_parallel_spawn) or sequentially (via individualsubplan_spawndecisions). Without asubplan_parallel_spawnwrapper, eachsubplan_spawndecision results in sequential execution. - Applicable invariants are enforced during Strategize by adding
invariant_enforceddecisions to the tree, which constrain downstream decisions and child plans. - The parent plan is responsible for merging results.
This is core to the long-term objective: tackling large tasks while only recomputing parts of the decision tree when corrected.
Hierarchical Decomposition for Scale
When handling massive tasks (e.g., converting Firefox to Rust), the system uses hierarchical decomposition. Each level spawns child plans (which are themselves full Plans with their own decision trees):
-
Root Plan: High-level architectural decisions and invariants
- "Convert Firefox Renderer to Rust"
- Invariant: "Maintain API compatibility with existing C++ callers"
- Decision: "Start with leaf modules, work inward"
- Context: Module dependency graph (2,847 modules)
-
Subsystem-Level Plans: Major component decisions (spawned via
subplan_parallel_spawnfor independent subsystems)- "Phase 1: Convert utility libraries (no external deps)"
- Each subsystem plan gets its own bounded context and inherits parent invariants
-
Module-Level Plans: Individual module conversions
- "Convert string_utils module"
- Context: Only the 47 functions and 12 dependent files
- Decision: "Use Rust's String type"
-
File-Level Plans: Specific file changes
- Actual code transformations
- Minimal context needed
At each level, only the relevant context is loaded. The persistent decision graph means we can always reconstruct why we're converting a particular module and what constraints (including inherited invariants) apply from higher-level decisions.
Child Plan Spawning Mechanism
In the actor definition for the execution actor, tool nodes can directly invoke registered tools to trigger child plans. The local/create-subplan tool is independently registered and referenced by name in the actor graph node. During execution, subplan_spawn decisions are realized as actual child plans, and subplan_parallel_spawn groups trigger concurrent spawning of all enclosed child plans.
# Example: Execution actor with subplan spawning capability.
# The graph uses a tool node referencing a named registered tool.
actors:
code_executor:
type: graph
config:
actor: anthropic/claude-3-opus
skills:
- local/plan-tools # Skill containing create_subplan for LLM tool-calling
- local/file-ops
routes:
execute_workflow:
nodes:
- name: spawn_test_subplan
type: tool
tool: local/create-subplan # Named tool from Tool Registry
The local/create-subplan tool is independently registered via its own YAML configuration file:
# File: tools/create-subplan.yaml cleveragents: version: "3.0"tool: name: local/create-subplan description: "Spawn a subplan for a given action" source: custom
input_schema: type: object properties: action: { type: string } target_files: { type: array, items: { type: string } } required: [action]
capability: writes: true checkpointable: false side_effects: [spawn_subplan]
code: | subplan = ctx.spawn_subplan( action=params["action"], target_files=params.get("target_files", []) ) return {"subplan_id": subplan.id}
The local/plan-tools skill references this tool (and others) by name:
# File: skills/plan-tools.yaml
skill:
name: local/plan-tools
description: "Tools for spawning and managing subplans"
tools:
- local/create-subplan
Child Plan Execution Modes
- Sequential: Individual
subplan_spawndecisions without asubplan_parallel_spawnwrapper execute one after another. If one fails, subsequent child plans are not started. - Parallel: Multiple
subplan_spawndecisions grouped under asubplan_parallel_spawndecision execute concurrently. If one fails, others can continue. Thesubplan_parallel_spawndecision acts as a container that signals the system to spawn all enclosed child plans simultaneously.
Child Plan Failure Handling
- Parallel execution (via
subplan_parallel_spawn): Other parallel child plans continue even if one fails. - Sequential execution: Subsequent child plans are not started if a prior one fails.
- Note: An "error" only occurs if an exception is thrown by the application (a bug). Plan failures (e.g., tests don't pass) are handled within the plan's logic, not as application errors.
Child Plan Result Merging
The way child plan results are merged depends on the resource type:
- Git-compatible resources (source code, text files): Git-style merge
- Databases: Transaction coordination or sequential application
- Other resources: Pluggable merge strategies based on resource type
- Non-mergeable resources: May require sequential execution only
The Plan "Decision Tree" and Visualization
CleverAgents intends to record enough information to render:
- an ASCII tree in the TUI, and
- optionally a GUI tree via visualization tools (D3/Cytoscape) once the data exists.
This implies each plan should persist:
- decisions made,
- the rationale (or at least the prompt/context snapshot that produced it),
- dependencies ("this decision influenced these child plans").
This is required for "correcting plans" (see Behavior section).
Decision Data Model
!!! adr "Architecture Decision" The decision data model, decision types, and decision tree semantics are defined in ADR-007: Decision Tree and Correction.
Relationship Between Plan Description and Decisions
Each plan has a description field (inherited from the action's description, potentially with argument substitutions). This description acts as the primary component of the prompt fed to the strategy actor during the Strategize phase.
Decisions are choices that are NOT explicitly defined by the plan description. They represent the gaps, ambiguities, or implementation details that must be resolved to execute the plan.
For example:
- Plan description: "Increase test coverage to 85%"
- Decisions that emerge:
- "Which modules should be prioritized?" (not specified in description)
- "Should we use mocks or integration tests for the database layer?" (not specified)
- "Should we refactor the auth module to make it more testable, or write tests around it as-is?" (not specified)
Decision Making Based on Autonomy Level
Who makes decisions depends on the plan's automation profile:
| Profile Flag | Who Makes Decisions |
|---|---|
auto_decisions_strategize = 1.0 (always manual) |
User is prompted for each decision point during Strategize |
auto_decisions_strategize = 0.0 (always automatic) |
Strategy actor makes decisions autonomously, records reasoning |
auto_decisions_execute = 1.0 (always manual) |
User is prompted for each decision point during Execute |
auto_decisions_execute = 0.0 (always automatic) |
Execution actor makes decisions autonomously |
Intermediate thresholds (e.g., auto_decisions_strategize = 0.7) cause the system to auto-decide only when the Semantic Escalation confidence score meets or exceeds the threshold; otherwise the user is prompted.
When automation allows automatic decisions, the strategy actor uses its best judgment based on context, and records its reasoning in the decision's rationale field.
When user input is required, the system pauses and prompts the user:
Decision required: Which modules should be prioritized for test coverage?Options identified by the strategy actor:
- auth module (currently 45% coverage, high risk)
- payment module (currently 52% coverage, high risk)
- user module (currently 71% coverage, medium risk)
Your choice (or provide custom guidance): _
The Prompt as the Root Decision
The prompt passed to the strategize actor is itself a decision node in the decision tree—specifically, it's the root decision of type prompt_definition.
This is important because:
-
Every plan has its own prompt: The root plan's prompt comes from the action description + user arguments. Child plan prompts are created by parent plans during their execution.
-
Parent plans create child plan prompts: When a parent plan spawns a child plan, it decides what prompt to give that child plan. This is recorded as a
prompt_definitiondecision in the parent's tree, and becomes the root decision of the child plan's tree. -
Invariants flow into the decision tree: During Strategize, applicable invariants (from global, project, action, and plan scopes) are reconciled via the Invariant Reconciliation Actor and recorded as
invariant_enforceddecisions, making them explicit constraints that influence downstream decisions and child plans. -
Unified correction mechanism: Since the prompt, invariants, and all other decisions are part of the same tree, correcting any of them uses the same
agents plan correctcommand.
Plan: 01KH29QDEE6DZTXKWNKCV8VP0F
├── [prompt_definition] "Increase test coverage to 85% for the whole project."
├── [invariant_enforced] "Prioritize all functionality related to financial transactions and user management"
├── [invariant_enforced] "All API calls over TCP must be mocked"
├── [strategy_choice] "Prioritize auth and payment modules, implement them in parallel before the rest"
├── [subplan_parallel_spawn] "Implement auth and payment modules, in parallel"
│ ├── [subplan_spawn] "Write tests for auth module"
│ │ └── Plan: 01KH29R8WPKPBHRY7Q0NA9XW86
│ │ ├── [prompt_definition] "Write unit tests for auth module using mocks for the remote API calls"
│ │ ├── [implementation_choice] "Test login flow first"
│ │ └── ...
│ └── [subplan_spawn] "Write tests for payment module"
│ └── Plan: 01KH29RN2YKSXMTBDG82AKRHRA
│ ├── [prompt_definition] "Write unit tests for payment module using"
│ └── ...
└── [subplan_parallel_spawn] "Write tests for all modules except the auth and payment modules"
└── ...
Correcting Decisions (Including Prompts)
All corrections use the same unified command:
agents plan correct <decision_id> --mode=<mode> --guidance "<corrected decision text>"
Parameters:
<decision_id>: The ULID of the decision to correct--mode: Eitherrevert(rollback and re-run) orappend(add fix at end)--guidance: Free-form text specifying what the correct decision should be
Examples:
# Correct a strategy choice agents plan correct 01ARZ3NDEKTSV4RRFFQ69G5FAV --mode=revert \ --guidance "Prioritize the payment module first, not auth, due to upcoming deadline"# Correct the root prompt to be more specific agents plan tree <plan_id> # Shows: [prompt_definition] id=01ARZ3NDEKTSV4RRFFQ69G5FAV "Increase test coverage to 85%"
agents plan correct 01ARZ3NDEKTSV4RRFFQ69G5FAV --mode=revert </span> --guidance "Increase test coverage to 85%, prioritizing auth and payment modules. Use mocks for database tests, not integration tests."
# Correct a subplan's prompt (originally created by parent plan) agents plan correct 01BRZ4PDFLUTW5SSGR70H6GBW --mode=revert </span> --guidance "Write unit tests for auth module, focusing on edge cases for token expiration"
# Append a fix rather than rewriting history agents plan correct 01ARZ3NDEKTSV4RRFFQ69G5FAV --mode=append </span> --guidance "The previous approach missed error handling tests - add comprehensive error path coverage"
# Remove an invariant that shouldn't apply agents plan correct 01CRZ5QEHMVUX6TTHR81I7HCX --mode=revert </span> --guidance "Remove this invariant - TCP mocking is not needed for this module since it has no network calls"
# Add a missing invariant to the plan agents invariant add --plan 01HXM8C2ZK4Q7C2B3F2R4VYV6J "All database queries must use parameterized statements"
Note: CLI commands should not require interactive input. The --guidance parameter provides the correction inline.
When to correct the prompt vs. a specific decision:
| Situation | Correction Approach |
|---|---|
| Original request was too vague | Correct the prompt_definition decision |
| Strategy actor made a bad choice on a specific question | Correct that specific decision |
| Parent plan gave a child plan a bad prompt | Correct the child plan's prompt_definition |
| An invariant should not apply to this plan | Correct (remove) the invariant_enforced decision |
| A missing constraint should be added | Add a new invariant_enforced decision via agents invariant add --plan or correct the plan's strategy |
| Sequential child plans should run in parallel | Correct the relevant decisions to use a subplan_parallel_spawn grouping |
Because the prompt is part of the decision tree, the system automatically knows that correcting it invalidates all downstream decisions in that plan (and its child plans).
Decisions are created during both the Strategize and Execute phases. The Strategize phase produces the initial decision tree — strategy choices, invariant enforcement, resource selections, and child plan blueprints. The Execute phase may create additional decisions constrained by those made during Strategize, reflecting discoveries and adjustments that arise during execution. If Execute finds that Strategize-phase constraints are too restrictive to proceed, the plan reverts to Strategize for adjustment (see Phase Reversion below). The decision tree captures what choices were made and why, enabling correction and replay.
Decision Record Structure
Decision: # Identity decision_id: ULID # Unique identifier plan_id: ULID # Parent plan this decision belongs to parent_decision_id: ULID | null # Parent decision (for tree structure) sequence_number: int # Order within the plan's decisions# Classification decision_type: enum - prompt_definition # The prompt/description for this plan (root decision) - invariant_enforced # An invariant (from global, project, action, or plan scope) applicable to this plan, added as a constraint - strategy_choice # High-level approach decision during Strategize - implementation_choice # How to implement a specific task - resource_selection # Which resources to read/modify - subplan_spawn # Decision to create a child plan (spawned later in Execute) - subplan_parallel_spawn # Decision to spawn a group of child plans in parallel (contains subplan_spawn children) - tool_invocation # Which skill/tool to use - error_recovery # How to handle a failure - validation_response # Response to validation failure - user_intervention # User provided guidance/correction
# The Decision Itself question: str # What question was being answered chosen_option: str # What was decided alternatives_considered: list[str] # Other options that were evaluated confidence_score: float | null # 0.0-1.0 if the actor provided confidence
# Context Snapshot (for replay) context_snapshot: hot_context_hash: str # Cryptographic hash of the exact context hot_context_ref: str # Pointer to the full stored snapshot relevant_resources: list[ResourceRef] # Every file/symbol that influenced this decision actor_state_ref: str # Complete LangGraph checkpoint
# When the system decides "refactor the authentication module to use async patterns," # it permanently records: # - Which files were examined to make that decision # - What symbols and dependencies were traced # - The exact code state that was analyzed # - The reasoning chain that led to this choice # - Alternative approaches that were considered but rejected
# Rationale rationale: str # Why this option was chosen actor_reasoning: str | null # Raw LLM reasoning if available
# Downstream Impact (populated during Execute phase) downstream_decision_ids: list[ULID] # Decisions that depend on this one downstream_plan_ids: list[ULID] # Child plans spawned because of this decision artifacts_produced: list[ArtifactRef] # Files/outputs created under this decision
# Timestamps created_at: datetime
# Correction Metadata is_correction: bool # Was this decision a correction of another? corrects_decision_id: ULID | null # If correction, which decision was replaced correction_reason: str | null # Why the correction was made superseded_by: ULID | null # If this decision was later corrected
Decision Timing
| Phase | Decision Activity |
|---|---|
| Strategize | Initial decisions are created, including invariant_enforced decisions for applicable invariants, strategy_choice decisions, subplan_spawn decisions, and subplan_parallel_spawn decisions grouping parallel child plans. downstream_plan_ids is empty. The system aims to make reasonable, comprehensive decisions at this stage to guide execution. |
| Execute | Child plans are spawned and downstream_plan_ids is populated based on subplan_spawn decisions. Additional decisions may be created during execution — these are constrained by Strategize-phase decisions and arise from discoveries made during execution (e.g., implementation_choice, resource_selection, tool_invocation, error_recovery). If Execute determines that Strategize-phase constraints are too restrictive, the plan reverts to Strategize for adjustment rather than overriding those constraints. |
| Apply | No new decisions. History can be flagged for cleanup after successful apply. |
Decision Tree Storage Schema
-- Core decision table CREATE TABLE decisions ( decision_id TEXT PRIMARY KEY, -- ULID plan_id TEXT NOT NULL, parent_decision_id TEXT, sequence_number INTEGER NOT NULL, decision_type TEXT NOT NULL, -- prompt_definition, invariant_enforced, strategy_choice, -- implementation_choice, resource_selection, subplan_spawn, -- subplan_parallel_spawn, tool_invocation, error_recovery, -- validation_response, user_intervention question TEXT, chosen_option TEXT NOT NULL, alternatives_considered TEXT, -- JSON array confidence_score REAL, rationale TEXT, actor_reasoning TEXT, context_snapshot TEXT NOT NULL, -- JSON blob is_correction BOOLEAN DEFAULT FALSE, corrects_decision_id TEXT, correction_reason TEXT, superseded_by TEXT, created_at TEXT NOT NULL,<span style="color: #5599ff; font-weight: 600;">FOREIGN</span> <span style="color: #5599ff; font-weight: 600;">KEY</span> (plan_id) <span style="color: #5599ff; font-weight: 600;">REFERENCES</span> plans(plan_id), <span style="color: #5599ff; font-weight: 600;">FOREIGN</span> <span style="color: #5599ff; font-weight: 600;">KEY</span> (parent_decision_id) <span style="color: #5599ff; font-weight: 600;">REFERENCES</span> decisions(decision_id), <span style="color: #5599ff; font-weight: 600;">FOREIGN</span> <span style="color: #5599ff; font-weight: 600;">KEY</span> (corrects_decision_id) <span style="color: #5599ff; font-weight: 600;">REFERENCES</span> decisions(decision_id), <span style="color: #5599ff; font-weight: 600;">FOREIGN</span> <span style="color: #5599ff; font-weight: 600;">KEY</span> (superseded_by) <span style="color: #5599ff; font-weight: 600;">REFERENCES</span> decisions(decision_id));
-- Downstream relationships (many-to-many for DAG) CREATE TABLE decision_dependencies ( upstream_decision_id TEXT NOT NULL, downstream_decision_id TEXT NOT NULL, dependency_type TEXT NOT NULL, -- 'decision', 'plan', 'artifact' downstream_ref TEXT NOT NULL, -- The actual ID of decision/plan/artifact
<span style="color: #5599ff; font-weight: 600;">PRIMARY</span> <span style="color: #5599ff; font-weight: 600;">KEY</span> (upstream_decision_id, downstream_decision_id, downstream_ref), <span style="color: #5599ff; font-weight: 600;">FOREIGN</span> <span style="color: #5599ff; font-weight: 600;">KEY</span> (upstream_decision_id) <span style="color: #5599ff; font-weight: 600;">REFERENCES</span> decisions(decision_id));
-- Correction history CREATE TABLE correction_attempts ( attempt_id TEXT PRIMARY KEY, -- ULID plan_id TEXT NOT NULL, original_decision_id TEXT NOT NULL, new_decision_id TEXT, original_subtree_snapshot TEXT, -- Reference to archived state correction_reason TEXT, status TEXT NOT NULL, -- 'pending', 'executing', 'completed', 'failed' created_at TEXT NOT NULL, completed_at TEXT,
<span style="color: #5599ff; font-weight: 600;">FOREIGN</span> <span style="color: #5599ff; font-weight: 600;">KEY</span> (plan_id) <span style="color: #5599ff; font-weight: 600;">REFERENCES</span> plans(plan_id), <span style="color: #5599ff; font-weight: 600;">FOREIGN</span> <span style="color: #5599ff; font-weight: 600;">KEY</span> (original_decision_id) <span style="color: #5599ff; font-weight: 600;">REFERENCES</span> decisions(decision_id), <span style="color: #5599ff; font-weight: 600;">FOREIGN</span> <span style="color: #5599ff; font-weight: 600;">KEY</span> (new_decision_id) <span style="color: #5599ff; font-weight: 600;">REFERENCES</span> decisions(decision_id)
);
Action
!!! adr "Architecture Decision" Actions as reusable plan templates and the action-to-plan transition are defined in ADR-006: Plan Lifecycle.
What an Action Is
An action is a reusable plan template that is not associated with any projects yet.
Actions are created via CLI commands with a required --config YAML file that fully defines the action. Any CLI options supplied alongside the config file act as optional overrides for values in the YAML.
Examples:
- "Increase test coverage to 80%"
- "Refactor module X to be async-safe"
- "Write an RFC for feature Y"
- "Provision an infra cluster and validate access" (non-code)
Actions are intentionally project-agnostic so they can be reused across projects.
An action is the first stage of a plan — before it is used. Because of this, invariants can be attached to actions. When an action is used (via agents plan use), any invariants attached to the action are carried forward as plan-level invariants on the resulting plan. This allows teams to bake constraints directly into reusable templates. Invariants can also be added to a plan after the action is used, via agents invariant add --plan or --invariant flags on agents plan use.
Action Creation (CLI)
Actions are created using the CLI. The action is fully defined by the YAML configuration file:
agents action create --config ./actions/code-coverage.yaml
The YAML file provides the complete action definition, including the name, strategy-actor, execution-actor, definition-of-done, arguments, invariants, and all other properties.
Required parameters:
--config: YAML configuration file that fully defines the action (name, strategy-actor, execution-actor, definition-of-done, etc.)
Arguments defined in the YAML's args section are values that will be:
- Injected into the description and/or definition of done (via templating)
- Passed into the context of the actors
- Required when using the action on projects
Action Data Model (Expanded)
A plan in the Action phase has:
1) name (namespaced)
Format:
[server:][namespace/]<name>
Rules:
- If server is omitted, default server is assumed unless namespace is
local. - If namespace is omitted, default is
local. - Names should be stable identifiers (kebab-case recommended).
Examples:
local/code-coveragemyusername/code-coveragemyorgname/code-coverageprod:myorgname/code-coverage(server-qualified)
2) short_description
Optional at creation; auto-filled if blank.
3) long_description
Optional but recommended for reusable actions.
4) definition_of_done (DoD)
Required. Must be explicit and testable.
5) actors
Two actors minimum, two optional:
- strategy_actor (planner/architect) — required
- execution_actor (builder/implementer) — required
- estimation_actor (cost/risk estimator) — optional
- invariant_actor (Invariant Reconciliation Actor) — optional; resolves invariant conflicts when the plan enters Strategize. Lookup falls back to project, then global config.
Actors can be:
- an LLM agent (built-in),
- a graph (custom yaml, or built-in),
Note: graphs are hierarchical allowing them to reference other actors as nodes.
Actor abstraction is central: an actor may be a single agent or an entire graph.
6) reusable (boolean)
- Default:
true. - If
true: using the action creates a new plan in Strategize while leaving the action available. - If
false: action self-deletes (or auto-archives) after first use.
7) read_only (boolean)
- Default:
false. - If
true: the plan must only use read-only tools (tools withread_only: truein their capability metadata) and must never modify resources (even in sandbox). - Read-only actions are still useful for "investigation reports," architecture reviews, or dry-run planning.
8) inputs_schema (recommended addition)
To make actions genuinely reusable, actions should declare their inputs:
- required args (e.g., target coverage percent),
- optional args (e.g., test framework),
- validation rules (types, bounds).
Example:
target_coverage_percent: integer 0–100
9) automation_profile
The resolved automation profile name for this plan (e.g., trusted, auto, local/careful-auto). Determined at plan use time using the profile precedence rules (plan > action > project > global). Once set, it is locked to the plan.
Strategy (Strategize Phase)
!!! adr "Architecture Decision" The Strategize phase, strategy actors, and decision generation are defined in ADR-006: Plan Lifecycle.
Using an Action (Transition to Strategize)
The use command transitions an Action into the Strategize phase by applying it to one or more projects:
# Basic usage agents plan use local/code-coverage my-api-service# Multiple projects agents plan use local/schema-update </span> api-service </span> web-frontend </span> mobile-app
# With action arguments agents plan use local/code-coverage </span> my-api-service </span> --arg target_coverage_percent=85 </span> --arg test_framework=pytest
# With explicit automation profile agents plan use local/deploy-action </span> staging-env </span> --automation-profile manual
# With invariants attached at use time agents plan use local/code-coverage </span> my-api-service </span> --arg target_coverage_percent=85 </span> --invariant "All API calls over TCP must be mocked" </span> --invariant "Do not modify the payments module"
Parameters:
<PROJECT>: Project to apply the action to (positional, can be repeated for multi-project plans)--arg: Action argument values (format:name=value)--automation-profile: Override automation profile for this plan--invariant: Invariant to attach to the created plan (can be repeated). These are added as plan-level invariants in addition to any inherited from the action, project, or global scope.
When the action is used:
- A new plan is created with a unique ULID
- The plan's automation profile is resolved (plan > action > project > global precedence)
- Any invariants from the action are carried forward as plan-level invariants, combined with any
--invariantflags provided - The plan enters the Strategize phase
- The Invariant Reconciliation Actor computes the effective invariant view (resolving conflicts using plan > project > global precedence)
- The
strategy_actorbegins analyzing the project(s)
What Strategize Does
When an action is used on projects, it becomes a plan in Strategize.
Strategize is:
- read-only, producing a plan of attack,
- responsible for gathering context from project resources,
- responsible for collecting applicable invariants (from global, project, action, and plan scopes), computing the effective invariant view via the Invariant Reconciliation Actor (applying plan > project > global precedence), and recording them as
invariant_enforceddecisions, - responsible for generating a strategy and child plan blueprint (using
subplan_spawnandsubplan_parallel_spawndecisions), - not allowed to execute child plans or modify resources.
This "architect vs coder" separation is explicitly described as a core motivation.
Resource-aware dependency analysis: During the Strategize phase, the strategy actor employs specialized mechanisms to compute precise dependency closures:
# Pseudocode of what happens inside a strategy actor def compute_closure_for_refactoring(target_module): closure = ResourceClosure()<span style="opacity: 0.7;"># Direct file dependencies</span> closure.add_files(find_imports(target_module)) closure.add_files(find_includes(target_module)) <span style="opacity: 0.7;"># Symbol dependencies</span> <span style="color: magenta; font-weight: 600;">for</span> symbol <span style="color: magenta; font-weight: 600;">in</span> extract_exported_symbols(target_module): closure.add_files(find_symbol_usage(symbol, scope=<span style="color: #66cc66;">'project'</span>)) <span style="opacity: 0.7;"># Test dependencies</span> closure.add_files(find_tests_for_module(target_module)) <span style="opacity: 0.7;"># Build system dependencies</span> closure.add_files(find_build_references(target_module)) <span style="color: magenta; font-weight: 600;">return</span> closure
The system leverages several key insights:
- Modular boundaries exist: Even in legacy codebases, there are natural boundaries
- Changes are incremental: We don't convert 50,000 files atomically
- Dependencies are sparse: Most modules depend on a small fraction of the codebase
- Interfaces are narrow: Public APIs are much smaller than implementations
Strategize Data Model
A plan in Strategize contains all Action fields plus:
1) projects
A list of projects the plan is used on.
Important: A strategy plan may target multiple projects. Multi-project work in one "window" is considered a major usability advantage over tools that require being run from a single directory.
2) strategy_context
A structured object describing:
- what resources were considered,
- how they were retrieved,
- what filtering/limits were applied,
- what the actor saw.
This matters because a plan must be debuggable and correctable later.
Recommended fields:
resource_refs: IDs of resources usedqueries: search queries performedselected_chunks: chunk IDs + sources + reasonsconstraints: context window limits, file ignore patternsgenerated_summaries: if summarization occurred
3) strategy
The output plan:
- steps (ordered and/or DAG),
- conditions/branches ("if tests fail, do X"),
- child plans to spawn (including which action templates to use, and whether they should run in parallel via
subplan_parallel_spawn), - evaluation criteria (how to know success),
- risk assessment.
4) execution_blueprint (recommended addition)
Strategize should output not only narrative text but also a machine-usable blueprint:
- list of tasks,
- required skills,
- expected outputs,
- dependencies between tasks.
This blueprint becomes the input to Execute.
5) cost_estimate and risk_estimate (optional)
Cost and risk estimation is optional but recommended for production use.
When enabled, a specialized estimation actor analyzes:
- The initial prompt/request
- The strategy produced by the Strategize phase
- Historical data from similar plans (if available)
And produces estimates for:
- LLM tokens/cost range
- Number of steps/child plans expected
- Expected risk of rollbacks
- Estimated execution time
Implementation: Similar to how there's a strategy_actor and execution_actor for each action, there can be an optional estimation_actor whose entire job is cost/risk estimation. This actor runs after Strategize completes (before Execute) and its output is informational only.
# Example: Action with estimation actor (defined in the YAML config file)
agents action create --config ./actions/expensive-refactor.yaml
This becomes critical in server/multi-user usage and cost controls.
Execution (Execute Phase)
!!! adr "Architecture Decision" The Execute phase, sandbox model, and execution safety mechanisms are defined in ADR-006: Plan Lifecycle and ADR-015: Sandbox and Checkpoint.
What Execute Does
Execute is where the plan actually performs work, but in a sandboxed environment that can later be reviewed and applied.
Key properties:
-
Work happens in a sandbox All file modifications, generated artifacts, and intermediate outputs live in an isolated "execution workspace" until Apply.
-
Execute may spawn child plans Child plan spawning is a first-class behavior of Execute: a parent plan can distribute work to child plans (sequentially via
subplan_spawnor concurrently viasubplan_parallel_spawn) and merge results. -
Execute must support checkpointing / rollback (when enabled) Checkpointable tools allow rolling back to a checkpoint ID to recover from partial failure or wrong turns.
-
Execute produces a "reviewable diff" Diff review sandbox is described as a differentiating feature: users can inspect changes before applying.
Execution Workspace / Sandbox Model (Detailed)
!!! adr "Architecture Decision" Sandbox isolation, lazy sandboxing, and resource-type-specific sandbox strategies are detailed in ADR-015: Sandbox and Checkpoint.
A sandbox isolates plan execution from the real project resources until Apply.
Key Sandbox Principles
-
Lazy Sandboxing: Resources are sandboxed only when accessed, not upfront.
- A project may have many resources (git repo + 10 databases + cloud accounts)
- A plan may only modify one resource
- Only accessed resources are sandboxed
- Efficient for large projects
-
Per-Plan Sandboxes: Each plan and child plan has its own sandbox containing only the resources it edits.
-
Resource-Defined Strategy: The sandbox strategy is defined on each resource, not by skills or globally.
-
Cleanup Behavior:
- Sandboxes are cleaned up before application exit when possible
- Abandoned sandboxes (from crashes, etc.) are cleaned up on next application run
- Completed plan sandboxes are cleaned or archived based on retention policy
Sandbox Implementation Strategies
Different resource types require different sandbox strategies:
| Resource Type | Strategy | Rollback Mechanism |
|---|---|---|
git-checkout |
git_worktree |
Git reset/checkout |
git |
none |
N/A (represents a repo instance — not directly sandboxable) |
fs-mount |
copy_on_write or overlay |
Restore from snapshot |
fs-directory |
copy_on_write |
Restore from snapshot |
| Custom database types | transaction_rollback |
Transaction rollback |
| Custom API types | none |
Often not sandboxable |
1. Git worktree / branch sandbox (preferred for code)
- Create a worktree or temporary branch
- All modifications are commits or staged changes
- Apply merges/cherry-picks
Pros: natural rollback, diff support, efficient Cons: requires git
2. Filesystem copy sandbox
- Copy project directory to a sandbox directory
- Execute modifies sandbox copy
- Apply syncs diff back
Pros: simple Cons: expensive for huge repos
3. Overlay filesystem sandbox
- Use overlayfs-style "copy-on-write" to avoid full copies
Pros: efficient Cons: more complex, OS-dependent
4. Transaction-based sandbox (for databases)
- Begin transaction at sandbox creation
- All operations within transaction
- Rollback on failure, commit on apply
Pros: native to databases Cons: long-running transactions can cause issues
5. No sandbox (for non-sandboxable resources)
- Some resources cannot be sandboxed (certain APIs, cloud services)
- User proceeds at their own risk
- Plan should warn about non-sandboxable resources
Multi-Resource Sandboxing
When a plan accesses multiple resources:
- Each resource gets its own sandbox (based on its defined strategy)
- Sandboxes are independent
- Apply commits each sandbox separately
- If any sandbox Apply fails, others may still succeed (partial apply)
Complete isolation during execution prevents compound errors: Each plan executes in its own sandbox, which means:
Plan A (refactoring auth module):
- Sandbox A1: Contains only auth/*.cpp, auth_tests/*.cpp
- Cannot see Plan B's intermediate states
- Cannot accidentally depend on Plan B's half-done work
Plan B (updating API endpoints):
- Sandbox B1: Contains only api/.cpp, api_tests/.cpp
- Makes changes assuming current auth interface
Protected from Plan A's intermediate refactoring
Hierarchical merge resolution: When child plans complete, the parent plan performs intelligent merging:
def merge_subplan_results(subplan_results): # Group by resource type by_resource = group_by_resource_type(subplan_results)<span style="opacity: 0.7;"># Apply resource-specific merge strategies</span> <span style="color: magenta; font-weight: 600;">for</span> resource_type, changes <span style="color: magenta; font-weight: 600;">in</span> by_resource: <span style="color: magenta; font-weight: 600;">if</span> resource_type == <span style="color: #66cc66;">'git-checkout'</span>: merge_git_changes(changes) <span style="opacity: 0.7;"># Three-way merge</span> <span style="color: magenta; font-weight: 600;">elif</span> resource_type == <span style="color: #66cc66;">'fs-mount'</span>: merge_fs_changes(changes) <span style="opacity: 0.7;"># Copy-on-write reconciliation</span> <span style="color: magenta; font-weight: 600;">elif</span> resource_type.startswith(<span style="color: #66cc66;">'database'</span>): merge_db_changes(changes) <span style="opacity: 0.7;"># Sequential application</span> <span style="opacity: 0.7;"># Validate merged state</span> run_integration_tests()
Execution Data Model
A plan in Execute contains:
1) execution_context
The context used for execution (often smaller/more tactical than strategy context).
2) execution_log
Structured timeline of:
- tool calls (with parent skill noted),
- actor calls,
- outputs,
- errors and retries,
- checkpoints created.
This log is essential for debugging.
3) artifacts
Outputs produced:
- changed files,
- generated files,
- reports,
- diagrams,
- test outputs,
- diffs.
4) sandbox_ref
Pointer to the sandbox location/state:
- path, branch name, workspace ID, container ID, etc.
5) checkpoint_graph (if enabled)
A record of checkpoints:
- checkpoint ID
- timestamp
- tool responsible (and its parent skill)
- resources affected
- rollback instructions / metadata
Checkpointing in Execute (Core Safety Mechanism)
!!! adr "Architecture Decision" Checkpointing, rollback, and transaction safety are defined in ADR-015: Sandbox and Checkpoint.
The intended user-level behavior is:
- "Give me a checkpoint ID."
- Perform additional operations.
- "Roll back to checkpoint X."
Not all tools can support this; checkpointing must be declared per tool (in its capability metadata).
Tool-level checkpointability
Each tool declares (via its capability metadata):
checkpointable: true|falsecheckpoint_scope: what granularity of rollback is supported (file, transaction, commit, snapshot)rollback_mechanism: how rollback occurs
Examples:
- File tools (from a file-ops skill): snapshot file states pre-modification
- Git tools (from a git-ops skill): create commit or stash; rollback is reset/checkout
- Shell/CLI tools (running inside a container): rollback by restoring filesystem snapshot or reloading base image state
It should be noted that checkpointing is easier when tool scope is constrained (e.g., "only files within a docker image + git").
Plan-level rollback policy
Plans should have an option:
rollback_enabled: true|false
If disabled, the plan may use more generic/unsafe tools with fewer restrictions (useful for low-stakes tasks).
Execution as Transactions (Recommended)
Execution should be treated like a transactional pipeline:
-
Each step either:
- commits a checkpoint on success, or
- rolls back to the previous checkpoint on failure.
This is explicitly motivated by "partial failure leaves codebase inconsistent" and the need for transaction rollback.
Tool-Based Resource Modification (Modern Architecture)
!!! adr "Architecture Decision" The tool-based resource modification approach and tool execution model are defined in ADR-011: Tool System.
IMPORTANT: CleverAgents does NOT parse LLM output to extract code. Instead, it uses the modern tool-based approach pioneered by Claude Code, Cursor, and Aider where:
- LLMs call tools directly (
edit_file(),write_file(),delete_file(), etc.) — tools provided by referenced skills - Tools operate on the sandbox - each tool invocation modifies sandbox state directly
- ChangeSet is built from tool invocations - not by parsing LLM text output
- Validation runs on sandbox state - after tools execute, not on parsed output
This architecture provides:
- Atomic operations: Each tool call is a discrete, trackable change
- No parsing ambiguity: Tools have structured parameters (path, content, etc.)
- Resource-agnostic: Same pattern works for files, databases, APIs, any resource type
- Safety by design: Tools run in sandbox with defined capabilities and restrictions
- MCP compatibility: Tools from MCP-based skills map directly to MCP tools for external integrations
How It Works
sequenceDiagram
participant LLM as LLM Agent
participant Router as Tool Router
participant Sandbox as Sandbox
participant CS as ChangeSet
participant Val as Validator
LLM->>Router: Tool call (with parameters)
Router->>Router: Validate parameters
Router->>Router: Enforce capability restrictions
Router->>Sandbox: Execute tool in sandbox
Sandbox->>Sandbox: Operate on sandboxed state
Sandbox->>Sandbox: Record invocation
Sandbox->>Sandbox: Create checkpoint (if needed)
Sandbox->>CS: Emit Change record
CS->>CS: Accumulate into ChangeSet
CS->>Val: Submit for validation
Val->>Val: Run validators on sandbox state
Val->>Val: Generate diff from ChangeSet
Val-->>LLM: Present for review before Apply
Built-in Resource Tools
CleverAgents provides these core tools (via built-in skills) for resource manipulation:
| Tool | Description | Creates Change? |
|---|---|---|
read_file(path) |
Read file contents | No |
write_file(path, content) |
Create/overwrite file | Yes |
edit_file(path, changes) |
Apply targeted edits | Yes |
delete_file(path) |
Remove file | Yes |
move_file(src, dst) |
Rename/move file | Yes |
create_directory(path) |
Create directory | Yes |
list_files(pattern) |
List files matching glob | No |
search_files(pattern, content) |
Search file contents | No |
get_file_info(path) |
Get file metadata | No |
Each built-in tool automatically:
- Operates within sandbox boundaries
- Records changes to the ChangeSet
- Validates parameters against project configuration
- Enforces deny-list patterns (
.git/,node_modules/, etc.)
Why Not Parse LLM Output?
The obsolete approach of parsing markdown code fences has fundamental problems:
- Ambiguity: Is text explanation or code? Where does one file end and another begin?
- Fragility: Models output varying formats; regex parsing is brittle
- Loss of semantics: You lose the intent (create vs modify vs delete)
- No atomicity: Can't rollback individual operations
- Resource-limited: Only works for files, not databases or other resources
The tool-based approach solves all of these by making each operation explicit, typed, and trackable.
Semantic Error Prevention
!!! adr "Architecture Decision" The multi-layer error prevention strategy and guardrail architecture are defined in ADR-018: Semantic Error Prevention.
CleverAgents provides multiple layers of proactive error prevention that catch semantic errors before they can propagate through the system.
Layer 1: Decision-time Validation During Strategize
Every decision includes semantic validation:
Decision: Refactor payment module to async
alternatives_considered:
- "Convert to async/await patterns" (chosen)
- "Use thread pool with channels" (rejected: doesn't integrate with async ecosystem)
- "Keep synchronous with timeout" (rejected: doesn't solve core latency issue)
confidence_score: 0.85
validation_performed:
- Checked all payment API consumers can handle async
- Verified database driver supports async operations
- Confirmed no regulatory requirement for sync processing
Layer 2: Execution-time Semantic Guards
The execution actor uses a tool node that references the independently registered local/validate-api-compat tool:
# Actor graph uses a named tool node for semantic validation
actors:
code_executor:
type: graph
skills:
- local/semantic-validators # Skill containing validation tools for LLM tool-calling
nodes:
- name: semantic_validator
type: tool
tool: local/validate-api-compat # Named tool from Tool Registry
The local/validate-api-compat tool is independently registered via its own YAML:
# File: tools/validate-api-compat.yaml cleveragents: version: "3.0"tool: name: local/validate-api-compat description: "Check for breaking API changes and attempt auto-migration" source: custom
capability: writes: true checkpointable: true
code: | # Not just syntax checking - semantic validation old_api = extract_api_signature(previous_version) new_api = extract_api_signature(current_version)
breaking_changes = find_breaking_changes(old_api, new_api) if breaking_changes: affected_consumers = find_api_consumers(breaking_changes) migration_plan = generate_migration(breaking_changes) if can_auto_migrate(affected_consumers, migration_plan): apply_migration(migration_plan) <span style="color: cyan; font-weight: 600;">else</span>: raise SemanticError( "Breaking API changes require manual review", changes=breaking_changes, affected=affected_consumers )
The local/semantic-validators skill references this tool by name:
# File: skills/semantic-validators.yaml
skill:
name: local/semantic-validators
description: "Semantic validation tools for API compatibility and code invariants"
tools:
- local/validate-api-compat
Layer 3: Invariant Enforcement
!!! adr "Architecture Decision" Invariant scoping, enforcement mechanisms, and inheritance rules are defined in ADR-016: Invariant System.
Invariants are named constraints that guide and constrain plan execution. They can be attached at four scopes, all managed through the unified agents invariant command:
- Global invariants: Apply to all plans across all projects. Added via
agents invariant add --global. - Project invariants: Apply to all plans targeting a specific project. Added via
agents invariant add --project. Can also be attached at creation time viaagents project create --invariant. - Action invariants: Attached to an action and carried forward as plan-level invariants when the action is used. Added via
agents invariant add --action. Can also be attached at creation time viaagents action create --invariant. - Plan invariants: Apply to a specific plan and its child plans. Added via
agents invariant add --plan. Can also be attached at creation time viaagents plan use --invariant.
Precedence and conflict resolution: When invariants from different scopes conflict, narrower scopes override broader scopes:
- Plan-level invariants override project-level and global-level invariants.
- Project-level invariants override global-level invariants.
Note: Action invariants are carried forward as plan-level invariants when the action is used (see Action invariants above), so they participate in precedence at the plan tier — there is no separate action tier in the precedence chain.
Conflict resolution is performed by the Invariant Reconciliation Actor — a dedicated actor responsible for comparing invariants across scopes, identifying conflicts, and producing the final effective invariant view for a plan. The Invariant Reconciliation Actor is set at three levels via --invariant-actor:
- Global config: Set via
agents config set actor.default.invariant <ACTOR>. Defines the default Invariant Reconciliation Actor used when neither the project nor the plan specifies one. - Project-level: Set via
--invariant-actoronagents project create. If a project defines an Invariant Reconciliation Actor, it is used to reconcile that project's invariants against global invariants. - Plan-level: Set via
--invariant-actoronagents action create(carried forward when the action is used) oragents plan use(which overrides whatever was set on the action). If a plan has an Invariant Reconciliation Actor, it reconciles invariants from all scopes (plan, project, and global) and produces the final effective view for that plan.
The lookup order is: plan → project → global config. The first Invariant Reconciliation Actor found is used.
Invariant view calculation: When an action is used and a plan enters the Strategize phase, the Invariant Reconciliation Actor computes the effective invariant view by:
- Collecting all invariants from global, project, plan, and action scopes.
- Identifying conflicts (invariants from different scopes that contradict each other).
- Applying precedence rules (plan > project > global) to resolve conflicts.
- Producing the final set of effective invariants.
Each effective invariant is then recorded as an invariant_enforced decision in the plan's decision tree. This makes invariants visible, auditable, and correctable through the standard decision correction mechanism.
Child plan inheritance: When a top-level plan spawns child plans, the parent's effective invariant view (already reconciled) is passed down to each child plan. Child plans do not re-run reconciliation — they inherit the parent's resolved view.
Correcting invariants: The correction mechanism for invariant_enforced decisions supports two operations:
- Remove: Remove an existing invariant from the plan's decision tree (the invariant remains defined at its scope but is no longer enforced for this plan).
- Add: Add a new invariant to the plan. When adding, the user can select from invariants already accessible to the plan (those defined at the plan, action, project, or global scope), or provide free-form text to create a new ad-hoc invariant.
# Add invariants at different scopes agents invariant add --global "All public APIs must maintain backward compatibility" agents invariant add --global "Payment processing must be idempotent" agents invariant add --project local/api-service \ "Database transactions must complete within 5 seconds" agents invariant add --project local/api-service \ "Authentication must always use OAuth2" agents invariant add --plan <PLAN_ID> "All API calls over TCP must be mocked" agents invariant add --action local/code-coverage "Test files must not import production secrets"# List invariants agents invariant list --global agents invariant list --project local/api-service agents invariant list --plan <PLAN_ID> --effective # Shows reconciled view
# Remove an invariant agents invariant remove <INVARIANT_ID>
# Attach invariants at creation time (convenience) agents project create --invariant "All endpoints must validate auth tokens" local/api-service agents action create --config ./actions/code-coverage.yaml agents plan use local/code-coverage local/api-service --invariant "Mock all network calls"
# Correct an invariant decision (remove or replace via standard correction) agents plan correct <DECISION_ID> --mode=revert </span> --guidance "Remove this invariant - it does not apply to this module"
The system collects, reconciles, and checks invariants:
class InvariantEnforcer: def compute_effective_invariants(self, plan): """Compute the effective invariant view for a plan using the Invariant Reconciliation Actor.""" # 1. Collect raw invariants from all scopes raw = self.collect_all_invariants(plan)<span style="opacity: 0.7;"># 2. Find the Invariant Reconciliation Actor (plan -> project -> global config)</span> reconciler = ( self.get_plan_invariant_actor(plan) <span style="color: magenta; font-weight: 600;">or</span> self.get_project_invariant_actor(plan) <span style="color: magenta; font-weight: 600;">or</span> self.get_global_invariant_actor() ) <span style="opacity: 0.7;"># 3. Reconcile: apply precedence (plan > project > global), resolve conflicts</span> effective = reconciler.reconcile(raw, precedence=[<span style="color: #66cc66;">'plan'</span>, <span style="color: #66cc66;">'project'</span>, <span style="color: #66cc66;">'global'</span>]) <span style="color: magenta; font-weight: 600;">return</span> effective <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">collect_all_invariants</span>(self, plan): <span style="color: #66cc66;">"""Collect invariants from all scopes accessible to this plan."""</span> invariants = [] invariants.extend(self.get_global_invariants()) <span style="color: magenta; font-weight: 600;">for</span> project <span style="color: magenta; font-weight: 600;">in</span> plan.projects: invariants.extend(self.get_project_invariants(project)) invariants.extend(self.get_action_invariants(plan.action)) invariants.extend(self.get_plan_invariants(plan)) <span style="color: magenta; font-weight: 600;">return</span> invariants <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">check_invariant_preservation</span>(self, changes, enforced_invariants): <span style="color: #66cc66;">"""Check that changes respect all enforced invariants."""</span> <span style="color: magenta; font-weight: 600;">for</span> invariant <span style="color: magenta; font-weight: 600;">in</span> enforced_invariants: <span style="color: magenta; font-weight: 600;">if</span> <span style="color: magenta; font-weight: 600;">not</span> self.verify_invariant(invariant, changes): <span style="color: magenta; font-weight: 600;">return</span> InvariantViolation(invariant, changes) <span style="color: magenta; font-weight: 600;">return</span> Success()
Layer 4: Predictive Error Prevention
The system learns from past failures:
Error Pattern Database:
- pattern: "Async conversion in payment module"
historical_failures:
- "Race condition in payment confirmation"
- "Timeout handling breaks idempotency"
preventive_checks:
- "Add explicit transaction boundaries"
- "Verify idempotency keys are preserved"
- "Check distributed lock acquisition"
Apply Phase
!!! adr "Architecture Decision" The Apply phase, conflict resolution, and apply strategies are defined in ADR-006: Plan Lifecycle.
What Apply Does
Apply takes the sandboxed work product and makes it "real" in the project.
Core properties:
-
Apply is a controlled commit step Apply exists specifically to separate "generated work" from "committed work," enabling review and safer automation.
-
Apply is often the highest-risk step It changes real systems. This is where approvals and checks matter most.
-
Apply produces a terminal 'applied' state After successful apply, the plan reaches the
appliedterminal state within the Apply phase.
Apply Responsibilities (Recommended Checklist)
Apply should perform (configurable) validations before committing:
-
Diff review gate
-
If the
auto_applythreshold is not met (i.e., confidence < threshold, or threshold is1.0) in the automation profile, show:- changed files summary,
- full diff,
- risk warnings.
-
-
Conflict resolution
- If applying to a git repo, handle rebase/merge conflicts safely.
-
Audit log
- Record who applied, what changed, when, and why.
Validation in Apply
Validation runs during Execute, not during Apply. By the time a plan reaches the Apply phase, all required validations have already passed. Apply commits the sandbox changes to the real resources and does not re-run validations.
For full details on validation — including the Validation type system, modes, attachment scoping, failure handling, fix-then-revalidate loops, and data model — see the Validation section under Core Concepts.
Apply Data Model
A plan in Apply includes:
apply_summaryapplied_artifacts(final commit hash, merged PR link, file list)final_validation_results(per-validation tool invocation results:passed,message, anddatafor each)approval_record(if human approvals are required)deployment_record(optional, if apply triggers deploy)
Apply Terminal States
Apply is the final phase. It does not transition to another phase — instead, the plan reaches one of several terminal states within the Apply phase:
When Apply succeeds:
- plan.phase =
apply - plan.state =
applied - the sandbox may be cleaned up or archived depending on retention policy
- changes have been committed to real project resources
When Apply cannot complete within constraints:
- plan.phase =
apply - plan.state =
constrained - the Apply phase has determined that the current strategy's constraints prevent successful application (e.g., merge conflicts that violate invariants, validation failures that cannot be resolved within the strategy's scope, resource state drift that contradicts strategy assumptions)
- depending on the
auto_reversion_from_applyautomation profile flag, this may automatically trigger a reversion to Strategize, or the system may pause for user decision - sandbox remains intact pending reversion or user action
When Apply fails:
- plan.phase =
apply - plan.state =
errored - sandbox remains intact for inspection/retry
When Apply is cancelled:
- plan.phase =
apply - plan.state =
cancelled - sandbox remains intact for inspection
Project
!!! adr "Architecture Decision" The project concept, resource linking, and project-level configuration are defined in ADR-009: Project Model.
A project is the boundary that answers:
- "Where is the work happening?"
- "What can this plan read and write?"
- "What skills (and their tools) can this plan use?"
- "What context is available?"
A project is a collection of linked resources and configuration. Projects link to independently registered resources from the Resource Registry — they do not define resources inline. A resource can be linked to multiple projects, enabling shared resources across teams and workflows.
Important: Projects are created via CLI commands, NOT YAML configuration files.
Project Types: Local vs Remote
Projects are classified based on their resources:
| Type | Definition | Where Plans Can Execute |
|---|---|---|
| Local | Contains at least one local-only resource | Client only |
| Remote | All resources are remotely accessible | Client or Server |
This distinction matters for server mode: the server can only execute plans on remote projects because it needs network access to all resources.
Project Creation (CLI)
# Step 1: Register resources independently agents resource add git-checkout local/api-repo \ --path /repos/api-service \ --branch mainagents resource add local/database local/staging-db </span> --connection-string "postgresql://staging.example.com/mydb" </span> --read-only
# Step 2: Create the project agents project create "my-api-service"
# Step 3: Link resources to the project agents project link-resource "my-api-service" local/api-repo
agents project link-resource "my-api-service" local/staging-db --read-only
Resources are registered once and can be linked to multiple projects. The resource's type, sandbox strategy, and capabilities are defined by its resource type in the Resource Registry — not by the project.
Project Data Model
A project includes:
1) Identity
name— the namespaced project name (e.g.,local/api-service). This serves as the project's globally unique identifier; no separate ULID or ID is generated. The namespacing scheme ([[server:]namespace/]name) ensures global uniqueness across systems.namespace(follows same rules as actors:local/,<username>/,<orgname>/)is_remote(boolean, derived from resources)
2) Linked Resources
Resources are the "things you can act on." Projects link to independently registered resources from the Resource Registry rather than defining them inline. This means:
- A resource can be linked to multiple projects (shared resources).
- The resource's type, capabilities, sandbox strategy, and DAG relationships are defined by the resource itself — not by the project.
- Projects can apply project-level overrides when linking (e.g., marking a writable resource as read-only within a specific project context).
Each linked resource reference has:
resource_id(ULID reference to a Resource Registry entry). Resources can be referenced by name or ULID in CLI commands; name is resolved toresource_idat command time.project_read_only(boolean — project-level read-only override)alias(optional short name for referencing within the project)
The full resource details (type, location, sandbox strategy, capabilities, parent/child DAG, etc.) are stored in the Resource Registry. See the Resources section for details.
3) Context configuration
Project-level defaults:
- ignore patterns (like
.gitignoresemantics) - max file size
- indexing strategy
- preferred chunking/summarization policy (even if evolving)
- context retention policy
Multi-Project Operations
A single plan may target multiple projects (e.g., updating shared schemas across services). This is considered a key UX advantage over "run in one directory" systems. Because resources are independently registered and can be linked to multiple projects, shared resources across projects are a natural part of the architecture.
In multi-project execution:
- Strategize must clarify which steps affect which projects.
- Execution must isolate sandboxes per project OR define a composite sandbox. When multiple projects share the same resource, a single sandbox for that resource is used.
- Apply must commit changes to each project separately, with separate approval records if necessary.
- Tool resource bindings are resolved per-project — the same tool may bind to different resources depending on which project context it runs in.
Namespaces
!!! adr "Architecture Decision" The namespace system, resolution rules, and ownership model are defined in ADR-002: Namespace System.
Namespaces define ownership, scoping, and discoverability of actors, tools, skills, resources, resource types, actions, projects, and plans.
All named entities use the format <namespace>/<name>.
Namespace Types
| Namespace | Scope | Storage | Examples |
|---|---|---|---|
local/ |
Current machine only | Local database | local/my-reviewer, local/test-action |
<username>/ |
Personal server namespace | Server database | freemo/code-analyzer, jsmith/deploy-script |
<orgname>/ |
Organization namespace | Server database | cleverthis/standard-review, acme/deploy-action |
openai/, anthropic/, etc. |
Built-in LLM actors | N/A (built-in) | openai/gpt-4, anthropic/claude-3-opus |
Namespace Rules
-
local/- Reserved namespace for local-only items
- Exists only on the current machine
- Stored in local database
- Fast iteration, no sharing
- Default namespace when none specified
-
<username>/(e.g.,freemo/,jsmith/)- Personal namespace on the server
- Created when user registers an account
- Stored on server, synced when connected
- Used for reusable entities (actions, actors, tools, skills, resources) a user wants across machines
- Only the owning user can create/modify items
-
<orgname>/(e.g.,cleverthis/,acme/)- Organization namespace on the server
- Created when organization is registered
- Shared across team members
- All namespaced entities (actions, actors, tools, skills, resources, projects) can be centrally managed
-
Built-in Provider Namespaces (
openai/,anthropic/,google/, etc.)- Reserved for built-in LLM actors
- Automatically available when API keys are configured
- In server mode: available if logged in and server has keys
- In local mode: requires environment variables or app configuration
- Cannot be used for custom actors
Server-qualified Names
To disambiguate between servers (when connected to multiple):
dev:freemo/code-coverage(personal namespace on dev server)prod:cleverthis/deploy-action(org namespace on prod server)
This enables a pattern where:
- local machine runs a lightweight client
- server stores canonical definitions
- multiple servers can coexist
Actor
!!! adr "Architecture Decision" The actor abstraction, actor types, and actor graph composition are defined in ADR-010: Actor and Agent Architecture.
What an Actor Is
An actor is the abstraction that generalizes "agent" into "anything conversational."
- It can be as small as a single LLM agent.
- It can also be an entire graph that itself calls other actors/tools.
- Actors can be nested/hierarchical, enabling "orchestrator of orchestrators."
Every custom actor IS a graph (a LangGraph defined via YAML configuration). Even a simple actor wrapping a single LLM is technically a graph with one node.
Actor Naming
Actors are always named using <namespace>/<name> format:
local/my-reviewer- Local actorfreemo/code-analyzer- Personal server actorcleverthis/deploy-specialist- Organization actoropenai/gpt-4- Built-in LLM actor
Actor Definition (YAML Configuration)
Actors are defined via YAML configuration files. Tools and skills are also defined via their own YAML configuration files. YAML configuration is used for actors, tools, and skills (not for actions or projects).
Example actor configuration (see examples/ directory for full examples):
name: local/my-workflowcleveragents: version: "3.0" default_actor: workflow_controller
actors: # Simple LLM actor with skills referenced by name my_assistant: type: llm config: actor: openai/gpt-4 # Reference to built-in actor temperature: 0.7 system_prompt: | You are a helpful assistant. Current task: {{ context.task_description }} skills: - local/file-ops # Grants access to all tools in this skill - local/git-ops
# LLM actor with a composite skill (includes many sub-skills) data_processor: type: llm config: actor: anthropic/claude-3-opus system_prompt: | You are a data processing assistant. skills: - local/data-toolkit # A skill containing analysis + transformation tools
# Actor referencing another actor reviewer: type: llm config: actor: local/code-reviewer # Reference to another custom actor memory_enabled: true max_history: 20 skills: - local/file-ops - local/git-ops
routes: main_workflow: type: graph entry_point: start nodes: - name: analyze type: agent agent: my_assistant - name: process type: agent agent: data_processor edges: - source: start target: analyze - source: analyze target: process - source: process target: end
context: global: task_description: "Default task"
Actor Arguments
All actors can receive arguments when invoked, including built-in actors. Arguments are passed when:
- An action is used on projects (arguments flow to strategy/execution actors)
- An actor is directly invoked
Arguments are injected into the actor's context and can be used in Jinja2 templates within prompts.
For built-in actors (like openai/gpt-4), common arguments include:
temperaturemax_tokenssystem_prompt
Actor Composition (Hierarchical References)
Actors can reference other actors by name:
actors:
complex_workflow:
type: llm
config:
actor: local/base-analyzer # References another actor
Load order matters: Referenced actors must be loaded/defined before actors that depend on them.
This enables hierarchical composition where:
- Actor A's graph can include nodes that call Actor B (by name)
- Actor B itself is a graph that might call Actor C
- And so on...
Actor vs Agent (Relationship)
-
Agent: an actor that is specifically an LLM with tools and reasoning behaviors.
-
Actor: may be an agent, but may also be:
- a composite workflow,
- a multi-step graph,
- a wrapper around a third-party system (as long as it's "text in → text out" conversationally).
Actor Definition Fields (From Notes + Extended)
A robust actor schema should include:
name(namespaced, required in the YAML config file — defines the actor's registered identity)provider(LLM provider or runtime target)modelsystem_prompt(or prompt template)skill_access_policy(which skill categories/names are allowed or denied)graph_descriptor(for composite actors)memory_policy(per-plan/per-actor—see memory section)context_view_policy(what context this actor sees — includes ACMS settings:preferred_strategies,default_breadth,default_depth,depth_gradient,skeleton_budget_ratio,temporal_scope,auto_refresh,refresh_threshold)limits(token limits, tool call limits, retries)cost_policy(caps, budgets)metadata(use cases, version)
Actor Composition and Graphs
Actors can reference:
- other actors (by namespaced name)
- skills (by namespaced name — all tools from referenced skills become available)
- subgraphs
This is central to enabling both:
- multi-agent orchestration, and
- modular reuse of workflows.
Nodes in the Graph: Actors and Tools
Graph nodes can be any of:
- an actor (another LLM agent or composite workflow, referenced by name),
- a tool node — a deterministic, non-LLM step that directly invokes a tool. Tool nodes can:
- Reference a named registered tool by its fully-qualified name (e.g.,
tool: local/run-migrations). The tool must be registered in the Tool Registry viaagents tool add. Metadata can optionally be overridden at the point of use. - Define an anonymous inline tool using the same format as a tool YAML body (with
anonymous: true). This is useful for one-off, workflow-specific operations that don't warrant separate registration.
- Reference a named registered tool by its fully-qualified name (e.g.,
This is a powerful simplification: actors provide intelligence, tools provide capability (both as graph nodes and through skills for LLM tool-calling), and skills organize tools into reusable collections. Everything participates in the same graph.
Tool node with named tool reference:
nodes:
- name: run_db_migrate
type: tool
tool: local/run-migrations # Named tool from Tool Registry
override: # Optional metadata override
capability:
human_approval_required: true
name: spawn_tests type: tool tool: local/create-subplan # Another named tool
Tool node with anonymous inline tool:
nodes:
- name: custom_validation
type: tool
anonymous: true
description: "Validate output format before proceeding"
input_schema:
type: object
properties:
data: { type: object }
capability:
read_only: true
code: |
# Inline Python — same format as a named tool YAML body
if not params["data"].get("status"):
raise ValueError("Missing status field")
return {"valid": True}
Agent
!!! adr "Architecture Decision" The agent specialization and its relationship to the actor abstraction are defined in ADR-010: Actor and Agent Architecture.
Agent Definition
In CleverAgents, an agent is a specialized actor with:
- a conversational interface,
- tool-calling capability,
- potentially memory, planning heuristics, and role identity.
Examples of agent roles:
- planner/architect (strategy actor)
- coder/implementer (execution actor)
- reviewer/qa agent
- release/apply agent
The transcript explicitly discusses role separation like planner/coder/reviewer in context views/memory proposals.
Agent Behavior Configuration
Agents should be configurable without code changes:
- prompt templates
- tool sets
- safety constraints
- style constraints (verbosity, code style)
- reliability controls (self-checks, validations)
A design goal is user empowerment: "users customize LLM behavior without modifying core code."
Tools
!!! adr "Architecture Decision" The tool system, tool adapter layer, and tool lifecycle are defined in ADR-011: Tool System.
What a Tool Is
A tool is a namespaced, independently registered, callable operation. It is the atomic unit of execution in CleverAgents — the smallest piece of functionality that can read, write, or transform resources. Tools are defined in their own YAML configuration files, managed through the agents tool CLI commands, and registered in the Tool Registry.
Tools follow the same <namespace>/<name> naming convention as actors, skills, and other entities (e.g., local/run-migrations, cleverthis/validate-api, local/create-subplan). They support optional server-qualified prefixes for multi-server disambiguation (e.g., dev:freemo/custom-analysis).
Validation as a Tool subtype: A Validation is a specialized subtype of Tool that extends the Tool class with validation-specific metadata (mode: required/informational) and a structured JSON return format ({ "passed": bool, "message": string, "data": object }). Because Validation extends Tool via standard class inheritance, it inherits all base properties and behaviors — registration, resource bindings, lifecycle hooks, capability metadata, source types (custom, MCP, agent_skill, builtin, and the validation-specific wrapped source for Validations that delegate to an existing Tool), and the ability to exist as a tool node in an actor graph. A Validation can also wrap an existing plain Tool via the wraps field, reusing the Tool's implementation and interpreting its output through a transform function — see the Tool Wrapping subsection under Core Concepts > Validation for details.
Key constraints on Validations:
- Always read-only: Validations observe and report but never modify resources.
writesis alwaysfalseandcheckpointableis alwaysfalse(read-only operations need no checkpointing). Theagents validation addcommand does not accept--writes/--no-writesor--checkpointable/--no-checkpointableflags. - Shared namespace: Validations and plain Tools share the same naming namespace in the Tool Registry. A Validation and a plain Tool cannot have the same name — attempting to register a Validation with a name already taken by a Tool (or vice versa) results in a conflict error.
- Superset/subset semantics: A Validation can be used anywhere a Tool is expected (since it IS a Tool), but a plain Tool cannot be used where a Validation is specifically required (e.g., attached to a resource as a validation).
- Unified management: Since Validations live in the Tool Registry, they are listed via
agents tool list(with--type validationto filter), inspected viaagents tool show, and removed viaagents tool remove. Only registration (agents validation add), attachment (agents validation attach), and detachment (agents validation detach) have validation-specific CLI commands.
Validations can be attached to resources directly, or to resources through projects or plans. The same mechanisms used to determine what tools can operate on what resources carry over to determining what validations apply to a particular resource. See the Validation section (immediately following this Tools section) for full details on the Validation type system, modes, attachment scoping, failure handling, and data model.
The Dual Role of Tools
Tools serve two distinct roles in CleverAgents:
-
As components of a Skill: A skill references tools by name to assemble a reusable capability collection. When an actor references a skill, all of that skill's tools (including those from included child skills) become available to the actor's LLM agent for tool-calling.
-
As tool nodes in an Actor graph: An actor's graph definition can include
type: toolnodes that directly invoke a specific tool. This is used for deterministic, non-LLM steps in a workflow — e.g., spawning a child plan, running validation, or executing a migration. The tool node either references a named registered tool or defines an anonymous inline tool.
block-beta
columns 3
space:3
block:header:3
A["Tool: Dual Role"]
end
space:3
block:role1:1
B["Role 1: In a Skill"]
C["Skill: local/devops"]
D["tools:"]
E[" - local/run-migrations"]
F[" - local/validate-schema"]
G["(tool-calling by LLM)"]
end
space:1
block:role2:1
H["Role 2: In an Actor Graph"]
I["Actor Graph node:"]
J[" name: run_db"]
K[" type: tool"]
L[" tool: local/run-migrations"]
M["(deterministic invoke)"]
end
Tool Configuration (YAML)
Tools are defined in their own YAML configuration files, separate from skills and actors. A tool YAML file declares the tool's identity, schema, capability metadata, and implementation:
# File: tools/run-migrations.yaml cleveragents: version: "3.0"tool: name: local/run-migrations description: "Run database migrations for the API service"
source: custom # mcp | agent_skill | builtin | custom
# Resource bindings — what resources this tool needs access to resources: db: type: local/database access: read_write required: true description: "Target database for migrations"
input_schema: type: object properties: direction: type: string enum: [up, down] count: type: integer default: 1 required: [direction]
capability: writes: true write_scope: resource_slots: [db] # References the "db" resource slot checkpointable: true checkpoint_scope: transaction side_effects: [schema_mutation]
code: | import subprocess direction = params["direction"] count = params.get("count", 1) db = ctx.resources["db"] # Access the bound database resource result = subprocess.run( ["alembic", direction, str(count)], capture_output=True, text=True, cwd=db.sandbox.root ) return {"stdout": result.stdout, "returncode": result.returncode}
Another example — a tool that wraps an MCP server endpoint:
# File: tools/create-github-issue.yaml cleveragents: version: "3.0"tool: name: local/create-github-issue description: "Create a GitHub issue via MCP"
source: mcp mcp_server: command: "npx @anthropic/mcp-github" env: GITHUB_TOKEN: "${GITHUB_TOKEN}" tool_name: create_issue # The tool name as exposed by the MCP server
capability: writes: true write_scope: [github:issues] checkpointable: false
And an Agent Skill tool:
# File: tools/deploy-staging.yaml cleveragents: version: "3.0"tool: name: local/deploy-staging description: "Deploy the current branch to the staging environment"
source: agent_skill agent_skill: path: ./skills/deploy-to-staging sandbox_policy: container allowed_tools: ["Bash(docker:)", "Bash(kubectl:)", "Read"]
capability: writes: true checkpointable: false side_effects: [deploy, infrastructure]
Tool Registration and Management
Tools are managed through the agents tool CLI commands:
# Register a new tool from its YAML configuration agents tool add --config ./tools/run-migrations.yaml# Update an existing tool (re-reads the config file, overwrites registration) agents tool add --config ./tools/run-migrations.yaml --update
# List all registered tools agents tool list
# Show details for a tool (schema, capability, references) agents tool show local/run-migrations
# Remove a tool agents tool remove local/run-migrations
Once registered, a tool is available to be referenced by skills (in their tools list) and by actor graphs (as type: tool nodes). Tools persist in the database (local or server) and follow the same namespace rules as actors and skills.
Anonymous Tools
An anonymous tool is an inline tool definition that appears directly in a skill YAML or an actor graph node. Anonymous tools use the same format as a named tool's YAML definition (same input_schema, capability, and code fields) but lack a namespaced name. They are:
- Not registered in the Tool Registry
- Not reusable — they exist only within the YAML file where they are defined
- Useful for one-off operations that are too specific to warrant separate registration
Anonymous tools in a skill YAML:
skill: name: local/my-skill tools: - local/run-migrations # Named tool reference - local/validate-api-compat # Named tool reference
anonymous_tools: # Inline definitions, same format as tool YAML - description: "One-off data cleanup for this project" input_schema: type: object properties: table: { type: string } capability: writes: true checkpointable: true code: | # ... Python code ... return {"cleaned": count}
Anonymous tools in an actor graph node:
nodes:
- name: custom_step
type: tool
anonymous: true
description: "Inline validation specific to this workflow"
input_schema:
type: object
properties:
data: { type: object }
capability:
read_only: true
code: |
# ... Python code ...
return {"valid": True}
The anonymous tool format is intentionally identical to the body of a named tool YAML — this means promoting an anonymous tool to a named, registered tool is a simple copy-paste into its own YAML file and agents tool add.
Metadata Overrides
When referencing a named tool in a skill or actor graph, its registered metadata can optionally be overridden at the point of use. This allows context-specific adjustments without modifying the tool's global registration.
Overriding tool metadata in a skill:
skill: name: local/strict-devops tools: - name: local/run-migrations override: capability: human_approval_required: true # Override: require approval in this skill write_scope: [database:staging] # Override: restrict scope for this context- local/validate-api-compat # No overrides, use as registered
Overriding tool metadata in an actor graph node:
nodes:
- name: safe_migrate
type: tool
tool: local/run-migrations
override:
capability:
human_approval_required: true
Overriding tool metadata when including a sub-skill:
When a skill includes another skill (importing all its tools), individual tools from the included skill can have their metadata overridden:
skill: name: local/production-ops includes: - name: local/devops-toolkit tool_overrides: - tool: local/run-migrations override: capability: human_approval_required: true # In this context, require human approval write_scope: [database:production]- <span style="color: cyan;">tool</span>: local/create-github-issue <span style="color: cyan; font-weight: 600;">override</span>: <span style="color: cyan; font-weight: 600;">capability</span>: <span style="color: cyan; font-weight: 600;">human_approval_required</span>: <span style="color: magenta; font-weight: 600;">true</span> <span style="opacity: 0.7;"># Require approval in this context</span>
Override rules:
- Overrides are shallow-merged — only the specified fields are replaced; unspecified fields retain their registered values.
- Overrides never persist back to the Tool Registry — they apply only at the point of use.
- Built-in tool metadata cannot be overridden (it is authoritative from the implementation).
- The override scope is limited to
capabilityanddescriptionfields. Schema (input_schema,output_schema) cannot be overridden because it would break callers' expectations.
Resource Bindings
!!! adr "Architecture Decision" Resource binding resolution, slot declarations, and project-specific binding are defined in ADR-008: Resource System and ADR-011: Tool System.
Tools operate on resources — git repositories, filesystems, databases, and more. The resource binding system declares and resolves the relationship between a tool and the resources it needs access to.
Resource Slots
A tool declares one or more resource slots in its YAML configuration. Each slot is a typed placeholder that specifies:
- Slot name: A logical name used to reference the resource within the tool's code and parameters.
- Resource type: The resource type required (e.g.,
git,fs-mount,local/database). The bound resource must be of this type (or a compatible subtype). - Access mode:
read_only,write_only, orread_write. - Description: Human-readable explanation of what the tool uses this resource for.
- Required/optional: Whether the slot must be bound for the tool to function.
Example tool YAML with resource slots:
tool: name: local/run-migrations description: "Run database migrations" source: customresources: db: type: local/database access: read_write required: true description: "Target database for migrations"
capability: writes: true write_scope: [db:migrations] # References the "db" slot checkpointable: true
code: | direction = params["direction"] db_resource = ctx.resources["db"] # Access the bound resource result = db_resource.handler.execute_migration(direction, db_resource.sandbox) return {"status": "ok"}
A tool that works with multiple resources:
tool: name: local/cross-repo-diff description: "Compare files across two git repositories" source: customresources: source_repo: type: git-checkout access: read_only required: true description: "Source repository to compare from" target_repo: type: git-checkout access: read_only required: true description: "Target repository to compare against"
capability: read_only: true
code: | source = ctx.resources["source_repo"] target = ctx.resources["target_repo"] # ... compare files across repos ...
Three Binding Modes
Resource slots are resolved to actual resources through one of three binding modes:
1. Contextual Binding (default)
The slot declares a resource type requirement, and the system resolves it from the plan's project context at activation time. This is the most common mode — the tool says "I need a git-checkout resource" and the system finds one among the project's linked resources.
resources:
repo:
type: git-checkout
access: read_write
# No `bind` field → contextual binding
Resolution rules for contextual binding:
- The system searches the plan's project for linked resources matching the slot's type.
- If exactly one resource of the right type exists, it is automatically bound.
- If multiple resources match, the system uses the slot name as a hint (e.g., a slot named
repoprefers a resource with aliasrepo). If ambiguous, the plan execution raises an error requiring explicit binding. - If no resource matches, the tool cannot be activated for this plan (a validation error is raised during plan creation).
2. Static Binding
The slot is hardcoded to a specific registered resource by name. This is useful for tools that always operate on the same resource, regardless of project context.
resources:
docs:
type: fs-mount
access: read_only
bind: local/company-docs # Static: always this resource
description: "Company documentation corpus"
Static bindings are resolved at registration time and validated — the named resource must exist and be of the correct type.
3. Parameter Binding
The resource reference is passed as a tool argument at invocation time. This is useful for tools that operate on user-specified resources.
resources: target: type: git-checkout access: read_only from_param: repository # Bound from the "repository" input parameter description: "Repository to analyze"
input_schema: type: object properties: repository: type: string description: "Name of the registered resource to analyze" required: [repository]
The from_param field links a resource slot to an input parameter. At invocation time, the system resolves the parameter value as a resource name from the Resource Registry and validates type compatibility.
Binding Resolution Flow
stateDiagram-v2
[*] --> ToolActivation: Actor references skill or tool node
state "For Each Resource Slot" as ForEach {
state binding_check <<choice>>
[*] --> binding_check: Check binding type
binding_check --> StaticBinding: has bind field
binding_check --> ContextualBinding: no bind, no from_param
binding_check --> ParameterBinding: has from_param
state "Static Binding" as StaticBinding {
[*] --> ResolveByName: Resolve from Resource Registry
ResolveByName --> ValidateType: Validate type compatibility
}
state "Contextual Binding" as ContextualBinding {
[*] --> SearchProject: Search plan's project resources
SearchProject --> FilterByType: Filter by resource type
state match_check <<choice>>
FilterByType --> match_check
match_check --> AutoBind: One match
match_check --> TryAlias: Multiple matches
match_check --> ValidationError: No matches
TryAlias --> AliasMatch: Try alias/name match
AliasMatch --> AutoBind: Match found
AliasMatch --> ValidationError: No match
}
state "Parameter Binding" as ParameterBinding {
[*] --> DeferBinding: Defer to invocation time
}
StaticBinding --> StoreBindings
AutoBind --> StoreBindings
ParameterBinding --> StoreBindings
state "Store in ToolActivationContext" as StoreBindings
}
state "Tool Invocation" as Invocation {
[*] --> ResolveParams: Resolve parameter-bound slots\nfrom invocation params
[*] --> ValidateAccess: Ensure sandbox exists\nValidate access mode
ResolveParams --> Execute
ValidateAccess --> Execute
state "Execute tool with bound resources" as Execute
}
ForEach --> Invocation
Invocation --> [*]
Built-in Tool Resource Bindings
Built-in tools have implicit resource slots that do not need to be declared in YAML (they are hardcoded in the implementation):
| Built-in Tool Group | Implicit Slot | Slot Type | Access |
|---|---|---|---|
file_operations (read_file, write_file, edit_file, etc.) |
directory |
fs-directory or git-checkout |
read_write |
directory_operations (create_directory, list_directory, etc.) |
directory |
fs-directory or git-checkout |
read_write |
search_operations (search_files, find_definition, etc.) |
directory |
fs-directory or git-checkout |
read_only |
git_operations (git_status, git_diff, git_log, etc.) |
repo |
git-checkout |
read_only |
Built-in tools accept both fs-directory and git-checkout types for file operations because a git-checkout resource's worktree root is an fs-directory. When binding to a git-checkout, the resource router automatically resolves to the worktree root fs-directory child for file operations. A standalone fs-mount resource also works — the router resolves through the fs-mount → root fs-directory chain.
Resource Discovery via Bindings
The binding system enables powerful resource discovery queries:
- "What tools can modify this resource?" → Find all tools with resource slots matching the resource's type and
access: read_writeoraccess: write_only. - "What tools can read this virtual file?" → Find the virtual file's physical children, then find tools with slots matching each physical resource's type.
- "What resources does this tool need?" → Inspect the tool's declared resource slots.
- "Is this tool compatible with this project?" → Check if the project's linked resources can satisfy all of the tool's required resource slots.
Tool Registry
CleverAgents maintains a Tool Registry — a persistent catalog of all independently registered tools:
@startuml
skinparam classAttributeIconSize 0
skinparam classFontSize 13
skinparam noteFontSize 11
skinparam defaultFontSize 12
class ToolRegistry {
- toolIndex : Map<String, ToolRecord>
--
+ add(config_path) : ToolRecord
+ update(name, config_path) : ToolRecord
+ remove(name) : void
+ lookup(name) : ToolRecord
+ list(filters) : ToolRecord[]
}
class ToolRecord {
+ name : String
+ description : String
+ source : String {mcp|agent_skill|builtin|custom}
+ config_path : String
+ input_schema : JSONSchema
+ output_schema : JSONSchema
+ capability_metadata : CapabilityMetadata
+ resource_slots : List<ResourceSlot>
+ code : String
}
ToolRegistry "1" *-- "0..*" ToolRecord : indexes >
note right of ToolRegistry
**Populated by:**
- agents tool add CLI command
- Dynamic refresh on MCP notifications
**Consumed by:**
- Skill registration
- Actor graph construction
- Resource binding resolution
- Plan validation
end note
@enduml
The Tool Registry works alongside the Skill Registry (described in the Skills section). Skills reference tools by name from the Tool Registry; the Skill Registry's flattened tool sets are composed from Tool Registry entries plus any anonymous inline tools.
Tool Interface and Architecture
Each individual tool — whether independently registered or defined as an anonymous inline tool — conforms to a uniform interface regardless of its source:
@startuml
skinparam classAttributeIconSize 0
skinparam classFontSize 13
skinparam defaultFontSize 12
skinparam linetype ortho
class Tool {
}
class Identity {
+ name : String
+ qualified_name : String
+ source : ToolSource
}
class Schema {
+ input_schema : JSONSchema
+ output_schema : JSONSchema
}
class CapabilityMetadata {
+ read_only : Boolean
+ writes : Boolean
+ write_scope : String
+ idempotent : Boolean
+ checkpointable : Boolean
+ side_effects : List<String>
+ cost_profile : String
+ human_approval_required : Boolean
}
class ResourceBindings {
+ slots : Map<String, ResourceSlot>
}
class ResourceSlot {
+ type : String
+ access : String
+ required : Boolean
+ bind : String
+ from_param : String
+ description : String
}
interface Lifecycle <<interface>> {
+ discover() : ToolDescriptor
+ activate() : void
+ execute(params, ctx) : Result
+ deactivate() : void
}
class ExecutionContext {
+ sandbox : Sandbox
+ plan : Plan
+ changes : List<Change>
+ resources : Map<String, BoundResource>
}
enum ToolSource {
mcp
agent_skill
builtin
custom
}
Tool *-- Identity
Tool *-- Schema
Tool *-- CapabilityMetadata
Tool *-- ResourceBindings
Tool *-- Lifecycle
Tool o-- ExecutionContext : uses at runtime >
ResourceBindings *-- "0..*" ResourceSlot
Identity --> ToolSource
@enduml
Every tool implements the same four lifecycle methods. The tool adapter layer is responsible for translating source-specific behavior into these methods.
Tool Adapter Layer
!!! adr "Architecture Decision" The adapter pattern for tool sources and the uniform tool interface are defined in ADR-011: Tool System.
Each tool source has a corresponding adapter that translates source-specific protocols into the uniform tool interface:
@startuml
skinparam classAttributeIconSize 0
skinparam classFontSize 13
skinparam defaultFontSize 12
interface "ToolInterface" as UTI <<interface>> {
+ discover() : ToolDescriptor
+ activate() : void
+ execute(params, ctx) : Result
+ deactivate() : void
}
class MCPToolAdapter {
+ discover() : ToolDescriptor
.. tools/list RPC → descriptors ..
+ activate() : void
.. spawn server, init JSON-RPC ..
+ execute(params, ctx) : Result
.. tools/call RPC → result ..
+ deactivate() : void
.. shutdown server ..
}
class AgentSkillAdapter {
+ discover() : ToolDescriptor
.. parse SKILL.md frontmatter ..
+ activate() : void
.. load SKILL.md body into agent context ..
+ execute(params, ctx) : Result
.. agent follows instructions, runs scripts ..
+ deactivate() : void
.. remove from context ..
}
class BuiltinAdapter {
+ discover() : ToolDescriptor
.. return hardcoded descriptors ..
+ activate() : void
.. no-op ..
+ execute(params, ctx) : Result
.. call native Python impl ..
+ deactivate() : void
.. no-op ..
}
MCPToolAdapter .up.|> UTI
AgentSkillAdapter .up.|> UTI
BuiltinAdapter .up.|> UTI
@enduml
MCPToolAdapter
Bridges external MCP servers into the tool model:
-
discover(): Spawns the MCP server process (or connects to a remote Streamable HTTP endpoint), performs the MCP
initializehandshake, negotiates capabilities, then callstools/listto enumerate available tools. Each MCP tool becomes a separateToolDescriptorwith itsinputSchemaand inferred capability metadata. -
activate(): Ensures the MCP server process is running and the JSON-RPC connection is healthy. For remote servers, validates the authentication token. Registers for
notifications/tools/list_changedso CleverAgents can dynamically update available tools. -
execute(): Translates a
Tool.execute(params, ctx)call into an MCPtools/callJSON-RPC request. Before dispatching:- Rewrites file paths to sandbox-relative paths
- Validates params against
inputSchema - Checks capability metadata against plan access policy
- Creates a checkpoint if the tool is marked checkpointable
After the MCP tool returns its
content[]response, the adapter:- Parses the result into the CleverAgents
Resultformat - Records any resource mutations as
Changeobjects in the plan'sChangeSet
-
deactivate(): Sends a clean shutdown to the MCP server process and closes the JSON-RPC connection.
Capability inference: MCP tools expose limited metadata (name, description, inputSchema). The adapter infers extended capability metadata using heuristics:
- Tools whose names contain
read,get,list,search,find→read_only: true - Tools whose names contain
write,create,update,delete,set→writes: true - All inferences can be overridden via the
overridesblock in tool or skill YAML
AgentSkillAdapter
Bridges Agent Skills Standard (SKILL.md folders) into the tool model. Agent Skills are fundamentally different from MCP tools — they are instruction-driven rather than schema-driven. An Agent Skill is not a single function call; it is a bundle of procedural knowledge that an LLM agent loads into its context and follows.
-
discover(): Scans the configured skill directory for a
SKILL.mdfile. Parses only the YAML frontmatter (name,description, optionalcompatibility,metadata,allowed-tools) to produce a lightweightToolDescriptor. This metadata is injected into the agent's system prompt in a structured format so the LLM can decide when the skill is relevant:<style="color: cyan; font-weight: 600;">available_agent_skills> <style="color: cyan; font-weight: 600;">agent_skill> <style="color: cyan; font-weight: 600;">name>deploy-to-staging</style="color: cyan; font-weight: 600;">name> <style="color: cyan; font-weight: 600;">description>Deploy the current branch to the staging environment.</style="color: cyan; font-weight: 600;">description> <style="color: cyan; font-weight: 600;">tool>local/deploy-staging</style="color: cyan; font-weight: 600;">tool> </style="color: cyan; font-weight: 600;">agent_skill> </style="color: cyan; font-weight: 600;">available_agent_skills>Discovery is low-cost — only ~50–100 tokens per Agent Skill for metadata. The full instructions are not loaded until activation.
-
activate(): When the LLM agent determines (or is instructed) that a task matches the skill's description, the adapter loads the full
SKILL.mdMarkdown body into the agent's active context. This injects step-by-step instructions, examples, edge cases, and references to bundled scripts. The tool is now "active" — the agent has the procedural knowledge to execute it.If the skill references additional files (
references/*.md,scripts/*.py,assets/*), these are made available on demand — the agent can read them as needed, following the Agent Skills Standard's progressive disclosure model. -
execute(): Unlike MCP tools (which are single RPC calls), Agent Skill execution is agent-mediated. The LLM agent follows the loaded instructions, potentially:
- Running bundled scripts via shell execution (sandboxed)
- Reading reference files for additional context
- Using other available tools (e.g., built-in file operations) as part of the procedure
- Making multiple tool calls in sequence to accomplish the workflow
The adapter wraps this execution in a tool execution context so that all mutations are tracked, sandboxed, and checkpointable. Script execution respects the
allowed_toolsandsandbox_policydeclared in the tool's YAML. -
deactivate(): Removes the skill's instructions from the agent's active context to free up token budget. The skill's metadata remains available for re-activation.
Key design principle: Agent Skills extend the agent's knowledge, not just its toolset. An Agent Skill can teach an agent a multi-step workflow that involves calling multiple other tools, making decisions based on intermediate results, and following domain-specific best practices — something a single MCP tool call cannot express.
BuiltinToolAdapter
Wraps CleverAgents' native resource operations as tools:
-
discover(): Returns hardcoded
ToolDescriptorobjects for each built-in operation. These descriptors have fully specified capability metadata since the implementation is first-party. -
activate(): No-op. Built-in tools are always available.
-
execute(): Calls the native Python implementation directly. Built-in tools operate through the resource abstraction layer, automatically integrating with sandbox path mapping, change tracking, and checkpointing.
-
deactivate(): No-op.
Built-in tool groups:
File Operations (file_operations):
read_file(path: str) -> str
write_file(path: str, content: str) -> None
edit_file(path: str, edits: list[Edit]) -> None
delete_file(path: str) -> None
move_file(source: str, destination: str) -> None
copy_file(source: str, destination: str) -> None
Directory Operations (directory_operations):
create_directory(path: str) -> None
list_directory(path: str, pattern: str = "*") -> list[str]
delete_directory(path: str, recursive: bool = False) -> None
Search Operations (search_operations):
search_files(pattern: str, content_pattern: str = None) -> list[Match]
find_definition(symbol: str) -> list[Location]
find_references(symbol: str) -> list[Location]
Git Operations (git_operations, when resource is a git repository):
git_status() -> GitStatus
git_diff(path: str = None) -> str
git_log(count: int = 10) -> list[Commit]
git_blame(path: str) -> list[BlameLine]
Each built-in tool:
- Has fully defined capability metadata
- Operates through the resource abstraction layer
- Automatically tracks changes to the ChangeSet
- Respects sandbox boundaries and deny-lists
Tool Capability Metadata (Critical for Safety)
MCP's metadata is not sufficient (read-only/idempotent is not enough; write scope is unclear). CleverAgents extends every tool — regardless of source — with a uniform capability metadata schema:
capability:
read_only: bool # Whether tool only performs read operations
writes: bool # Whether tool can modify resources
write_scope: # What the tool is allowed to mutate
- file_paths: ["src/**", "tests/**"] # Path patterns within bound resources
- resource_slots: ["repo", "db"] # Resource slot names (from resource bindings)
- environment: container | host
idempotent: bool # Whether repeated calls produce same result
checkpointable: bool # Whether tool supports checkpoint/rollback
checkpoint_scope: str # What can be rolled back (file, transaction, commit, snapshot)
side_effects: # Non-reversible effects
- install_packages
- mutate_infra
- send_email
cost_profile: # Usage constraints
rate_limit: "10/min"
estimated_cost: "$0.01/call"
human_approval_required: bool # Whether a human must approve invocation
Where metadata comes from per source:
| Source | Metadata origin | Override mechanism |
|---|---|---|
| Built-in | Hardcoded in implementation | Not overridable (authoritative) |
| MCP | Inferred from tool name/description + MCP annotations | overrides block in tool or skill YAML; override at skill/actor reference point |
| Agent Skill | Declared in SKILL.md frontmatter metadata + inferred from allowed-tools |
Tool YAML capability block; override at skill/actor reference point |
| Custom | Manually declared in tool YAML capability block |
override at skill/actor reference point (author is the source of truth for registered values) |
Read-Only Actions and Tool Access Control
When an action is marked read_only: true, it can only use tools that have read_only: true in their capability metadata. This is enforced at runtime by the tool execution context — any attempt to invoke a tool with writes: true from a read-only plan raises a AccessDeniedError.
Tool Execution Flow
When an LLM agent decides to use a tool (regardless of source), the following flow occurs through the unified execution pipeline:
1. LLM generates tool call
e.g., edit_file(path="src/main.py", changes=[...])
or: local/github.create_issue(title="Bug fix", body="...")
or: (activates Agent Skill "deploy-to-staging" via instructions)
↓
2. Tool Router receives call
- Resolves tool by name from the Tool Registry or actor's skill tool sets
- Validates parameters against inputSchema
- Checks capability metadata against plan's access policy:
• Is this tool in allowed skill categories?
• Does the plan allow writes?
• Is human approval required?
- If denied → return AccessDeniedError to LLM
↓
3. Resource Binding Resolution & Sandbox Context
- Resolve resource bindings for this tool:
• Static bindings: already resolved at registration
• Contextual bindings: resolve from plan's project resources
• Parameter bindings: resolve from invocation arguments
- Validate resource type compatibility for each slot
- Validate access mode (e.g., read_write tool on read_only resource → error)
- Ensure sandbox exists for each bound resource (lazy sandboxing)
- Maps logical paths to sandbox-relative paths via bound resource handlers
- Inject bound resources into ctx.resources[slot_name]
- If tool is checkpointable → create pre-execution checkpoint
↓
4. Adapter-Specific Execution
- MCP: sends tools/call JSON-RPC to server process
- Agent Skill: agent follows loaded SKILL.md instructions,
running scripts and tools in sandboxed shell
- Built-in: calls native Python implementation directly
- Custom: executes inline code with sandboxed context
↓
5. Change Recording
- If tool modified resources → create Change record(s)
- Append Change(s) to plan's ChangeSet
- Update sandbox state
- If checkpointable → record checkpoint for rollback
↓
6. Result Return
- Normalize result to uniform Result type
- Return to LLM agent for continued reasoning
Change Tracking from Tool Invocations
Critical Architecture Point: The ChangeSet is NOT built by parsing LLM output. It is built by recording the effects of tool invocations:
class ToolExecutionContext: """Context provided to every tool execution, regardless of source."""<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">__init__</span>(self, plan: Plan, sandbox: Sandbox, resources: <span style="color: cyan;">dict</span>[<span style="color: cyan;">str</span>, BoundResource]): self.plan = plan self.sandbox = sandbox self.resources = resources <span style="opacity: 0.7;"># slot_name → BoundResource</span> self.changes: <span style="color: cyan;">list</span>[Change] = [] <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">record_change</span>(self, change: Change) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Record a change made by a tool."""</span> self.changes.append(change) self.plan.changeset.add_change(change)class WriteFileTool: """Example: built-in tool for writing files."""
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">execute</span>(self, path: <span style="color: cyan;">str</span>, content: <span style="color: cyan;">str</span>, ctx: ToolExecutionContext) -> <span style="color: magenta; font-weight: 600;">None</span>: handler = ctx.sandbox.get_handler(path) change = handler.write(path, content, ctx.sandbox) ctx.record_change(change)
This approach means:
- Every resource modification is explicit and tracked
- The ChangeSet accurately reflects what was done, not what was said
- Rollback is precise (replay inverse of recorded changes)
- Audit logs show exactly which tool invocation produced each change
MCP Integration Details
!!! adr "Architecture Decision" The MCP integration strategy, concept mapping, and extended capabilities are defined in ADR-022: LangChain/LangGraph Integration.
MCP Concepts Mapping
| MCP Concept | CleverAgents Equivalent | Extension |
|---|---|---|
| Tool | Independently registered Tool (source: mcp), referenced by skills | Extended capability metadata (write_scope, checkpointable, side_effects) |
| Resource | Resource | Extended to support both read AND write operations |
| Prompt | Action template | Full plan lifecycle (Strategize → Execute → Apply) |
| Server | MCP Server connection (declared in tool or skill YAML) | Managed by MCPToolAdapter with lifecycle and reconnection |
| JSON-RPC | Internal adapter protocol | Abstracted behind Tool.execute() |
MCP Server Lifecycle Management
CleverAgents manages MCP server processes as part of the tool/skill/actor lifecycle:
sequenceDiagram
participant CLI as CLI
participant Reg as Registry
participant ActorRT as Actor Runtime
participant Adapter as MCPToolAdapter
participant MCP as MCP Server
note over CLI,Reg: Phase 1 - Registration
CLI->>Reg: agents tool add or skill add
Reg->>Reg: Validate server command or endpoint
Reg->>Reg: Store server config in record
note over ActorRT,MCP: Phase 2 - Actor Activation
ActorRT->>Adapter: Activate (for each MCP server)
Adapter->>MCP: Spawn process (stdio) or connect (HTTP)
Adapter->>MCP: MCP initialize handshake
MCP-->>Adapter: Capabilities
Adapter->>MCP: tools list
MCP-->>Adapter: Tool descriptors
Adapter->>MCP: Subscribe notifications tools list_changed
note over ActorRT,MCP: Phase 3 - Execution
ActorRT->>Adapter: LLM generates tool call
Adapter->>MCP: tools call (JSON-RPC)
MCP-->>Adapter: Result
Adapter->>ActorRT: Result + Change tracking
note over ActorRT,MCP: Phase 4 - Deactivation
ActorRT->>Adapter: Deactivate
Adapter->>MCP: Clean shutdown
Sandbox Path Rewriting for MCP Tools
MCP servers operate on real filesystem paths, but CleverAgents executes plans in sandboxes. The MCPToolAdapter transparently rewrites paths:
# Before sending to MCP server:
# Logical path: "src/main.py"
# Sandbox path: "/tmp/sandbox-01HXM/worktree/src/main.py"
# The adapter rewrites the tool arguments so the MCP server
# operates on sandboxed state without knowing about the sandbox.
This ensures MCP tools respect sandbox boundaries even though they have no awareness of the CleverAgents sandbox model.
Agent Skills Integration Details
Discovery and Progressive Disclosure
Agent Skills follow a three-tier progressive disclosure model that maps directly to the tool lifecycle:
| Tier | What loads | When | Token cost |
|---|---|---|---|
| Metadata | name + description from SKILL.md frontmatter |
Tool registration / actor activation | ~50–100 tokens per Agent Skill |
| Instructions | Full SKILL.md Markdown body | When LLM determines task matches the skill (activate phase) | Recommended < 5000 tokens |
| Resources | scripts/, references/, assets/ |
On demand during execution | Variable |
This means an actor can have dozens of Agent Skill tools available but only pay the token cost for their metadata at startup. Full instructions load only when relevant.
Agent Skills vs. MCP Tools: When to Use Which
| Dimension | MCP Tools | Agent Skills |
|---|---|---|
| Interaction model | Single function call with JSON params/result | Multi-step procedure with instructions the agent follows |
| Knowledge type | "Here is a function you can call" | "Here is how to accomplish a complex task" |
| Statefulness | Stateless per-call | Stateful across multiple tool calls within a procedure |
| Authoring | Implement an MCP server (code) | Write a SKILL.md file (prose + optional scripts) |
| Portability | Any MCP-compatible host | Any Agent Skills-compatible agent |
| Best for | Atomic operations (CRUD, queries, API calls) | Complex workflows (code review processes, deployment procedures, data analysis pipelines) |
Both can coexist in the same skill. A common pattern is an Agent Skill tool that teaches the agent a workflow which involves calling multiple MCP tools:
Agent Skill Tool: "local/deploy-staging"
SKILL.md instructions:
1. Run tests using the built-in shell tool
2. Create a PR using the GitHub MCP tool (create_pull_request)
3. Wait for CI using the GitHub MCP tool (get_check_runs)
4. Deploy using the AWS MCP tool (ecs_update_service)
5. Verify deployment using the HTTP MCP tool (fetch_url)
Tool-Level Checkpointability
Each tool declares whether it supports checkpoint/rollback via its capability metadata. Checkpointing behavior varies by source:
| Source | Checkpoint mechanism |
|---|---|
| Built-in file ops | Snapshot file state pre-modification; rollback restores file content |
| Built-in git ops | Create commit or stash; rollback via git reset / git checkout |
| MCP tools | Adapter-created checkpoints of affected resources; rollback replays inverse operations where possible |
| Agent Skills | Composite — each sub-tool-call within the skill's execution is individually checkpointed |
| Custom (containerized) | Filesystem snapshot or container image layer; rollback restores snapshot |
Checkpointing is easier when tool scope is constrained (e.g., "only files within a sandbox worktree + git"). The capability metadata's checkpoint_scope field communicates what granularity of rollback is supported.
When require_checkpoints is enabled on a plan, the plan may only use tools that have checkpointable: true. If disabled, the plan may use more generic/unsafe tools with fewer restrictions (useful for low-stakes tasks).
Validation
!!! adr "Architecture Decision" The validation system, gate types, and validation lifecycle are defined in ADR-013: Validation Abstraction.
What a Validation Is
A Validation is a specialized subtype of Tool designed to verify that work performed during plan execution meets specified quality, correctness, or compliance criteria. A Validation observes the current state of resources, evaluates a condition, and returns a structured pass/fail result. It never modifies resources — it only reads and reports.
Because Validation extends Tool via standard class inheritance, it is not a separate concept bolted onto the system — it is a tool, with additional constraints and metadata. This means validation benefits from the entire tool infrastructure: source types, resource bindings, lifecycle hooks, input/output schemas, registry management, skill composition, and actor graph integration. No parallel system is required.
Relationship to Tool
A Validation inherits all base Tool properties:
| Inherited from Tool | Additional in Validation |
|---|---|
name (namespaced) |
mode (required or informational) |
description |
Structured return format enforcement |
source (custom, mcp, agent_skill, builtin, wrapped) |
Read-only enforcement (writes: false, checkpointable: false) |
input_schema (JSON Schema) |
Attachment scoping (resource-centric, with optional project/plan scope) |
output_schema (JSON Schema) |
Attachment ULID (system-assigned per attachment) |
capability metadata |
wraps (optional reference to an existing Tool) |
resource_slots (resource bindings) |
transform (Python function to convert wrapped Tool output to validation format) |
lifecycle hooks (discover/activate/execute/deactivate) |
|
code (inline implementation) |
|
mcp_server / mcp_tool_name (MCP source) |
|
timeout |
|
idempotent |
Shared namespace: Validations and plain Tools share the same naming namespace in the Tool Registry. A name like local/run-tests can refer to either a Tool or a Validation — but never both simultaneously. Attempting to register a Validation with a name already taken by a plain Tool (or vice versa) results in a conflict error. This shared namespace enables unified management: agents tool list shows both types, agents tool show works for both, and agents tool remove removes either type.
Superset/subset semantics: Because Validation extends Tool, a Validation can be used anywhere a Tool is expected — it can appear in a skill's tool list, serve as a tool node in an actor graph, or be invoked directly by an LLM agent. However, a plain Tool cannot be used where a Validation is specifically required. For example, agents validation attach rejects a name that resolves to a plain Tool rather than a Validation. This distinction is enforced by the type discriminator stored in the Tool Registry: each entry is tagged as either tool or validation.
Read-Only Enforcement
Validations are always read-only. This is a hard constraint, not a convention:
writesis alwaysfalse. A Validation must never modify, create, or delete resources. It reads the current state and reports findings.checkpointableis alwaysfalse. Since a Validation produces no state changes, there is nothing to checkpoint or roll back.read_onlyis alwaystrue(implicit). This is the inverse ofwritesand is always set for validations.
This constraint is enforced at multiple levels:
- Registration time: The
agents validation addcommand does not accept--writes/--no-writesor--checkpointable/--no-checkpointableflags. Anywritesorcheckpointablevalues in the validation YAML configuration file are silently overridden tofalse. - Runtime enforcement: The execution runtime treats validation invocations as read-only operations. Resource bindings for validations are resolved with
access: read_onlysemantics — the validation receives a read-only view of the resource, not a writable handle. If a validation's implementation attempts to write (e.g., a custom Python validation calling a write API), the sandbox enforces the read-only constraint and the invocation fails. - Sandbox interaction: When a validation executes within a plan's sandbox, it reads the sandbox state (e.g., the modified files in a git worktree) but cannot modify it. This allows validations to verify the work-in-progress without risk of corrupting it.
The read-only guarantee has an important architectural consequence: validations are always safe to run. They cannot cause side effects, corrupt state, or require rollback. This enables aggressive parallelization, speculative validation (running validations before all work is complete to provide early feedback), and retry without concern for accumulated state changes.
Validation Mode
Each Validation has a mode that determines how the system responds to failure:
Required (mode: required):
- The validation must pass for execution to proceed to the Apply phase.
- When a required validation fails, the execution actor attempts to fix the underlying issue (e.g., fix failing tests, correct lint errors) within the bounds of the current strategy, then re-runs the validation.
- This fix-then-revalidate loop continues up to the configured retry limit.
- If self-fix is exhausted, the actor may request a strategy revision (re-running the Strategize phase for the affected subtree) or escalate to the user, depending on the automation profile.
- Required validations act as hard gates — they define "definition of done" for a plan.
Informational (mode: informational):
- The validation's result is recorded but does not block execution.
- When an informational validation fails, the result (including
messageanddata) is included in the plan's validation summary for human review, but the plan continues normally. - Informational validations are useful for advisory checks: bundle size reports, code complexity metrics, deprecation warnings, performance benchmarks, or any check where failure is noteworthy but not blocking.
- The distinction between required and informational allows teams to introduce new validations gradually — start as informational to measure impact, then promote to required once the codebase is clean.
The mode is set at registration time (in the validation YAML or via --required/--informational on agents validation add) and applies globally to that validation. The mode cannot be overridden per-attachment — if different scopes need different enforcement levels for the same check, register separate validations with different names and modes.
Structured Return Format
Every Validation must return a JSON object conforming to this structure:
{
"passed": true,
"message": "All 247 tests passed, 94% coverage",
"data": {
"tests_run": 247,
"tests_passed": 247,
"tests_failed": 0,
"coverage_percent": 94.2,
"coverage_threshold": 80,
"duration_seconds": 12.4
}
}
| Field | Type | Required | Description |
|---|---|---|---|
passed |
boolean | Yes | Whether the validation passed (true) or failed (false). This is the only field that drives system behavior (gating for required, recording for informational). |
message |
string | No | Human-readable summary of the result. Displayed in CLI output, plan summaries, and used by the execution actor to understand failures. Should be concise but actionable (e.g., "3 lint errors in src/api/handler.py" rather than just "failed"). |
data |
object | No | Arbitrary structured data in any format the validation chooses. This is the validation's primary output channel for rich, machine-readable results. Examples: test coverage reports, lint error lists with file/line/column, security vulnerability details, performance measurement data. The data field has no required schema — each validation defines its own structure. |
Return format enforcement: If a validation returns output that is not valid JSON, or valid JSON that lacks the passed boolean field, the system treats the invocation as an error (distinct from a validation failure). The validation result is recorded as "passed": false with a system-generated message indicating the malformed return, and the original output is preserved in data.raw_output for debugging.
The structured return format serves multiple consumers:
- The execution actor reads
messageanddatato understand failures and attempt fixes. A well-structureddatafield (e.g., with file paths and line numbers) enables more targeted fix attempts. - The plan summary records all validation results for human review and auditing.
- Downstream automation (CI scripts, dashboards) can parse
dataprogrammatically viaagents --format json plan show.
Attachment Scoping
Validations are not globally active — they must be explicitly attached to one or more scopes to take effect. This attachment model provides fine-grained control over which validations run for which work.
Attachment Model
A validation is always attached to a resource. The optional --project or --plan flag controls the scope under which the validation is active for that resource:
-
Direct attachment (
agents validation attach <RESOURCE> <VALIDATION> [args...])The validation is always active when that resource is accessed by any plan, regardless of which project or plan is involved. This scope is for invariants that must hold for a resource in all contexts — e.g., a lint check that must pass for a repository no matter what project is using it.
Use cases:
- Repository-wide lint/format checks
- Schema validation on database resources
- Security scanning on any codebase resource
- License compliance checks
-
Project-scoped attachment (
agents validation attach --project <PROJECT> <RESOURCE> <VALIDATION> [args...])The validation is active only for that resource when it is being interacted with through the specified project. Different projects using the same resource can have different validation requirements. This is the most common attachment scope.
Use cases:
- Unit test suites specific to a project
- Type checking for a TypeScript project (not relevant for a Python project using the same repo)
- API compatibility checks for a microservice project
- Bundle size checks for a frontend project
-
Plan-scoped attachment (
agents validation attach --plan <PLAN_ID> <RESOURCE> <VALIDATION> [args...])The validation is active only for that resource when it is being interacted with through the specified plan. This is the most targeted scope, useful for one-off or experimental validations.
Use cases:
- Temporary extra checks during a risky migration
- One-time regression validation for a specific bug fix
- Experimental validations being tested before promoting to project scope
- User-added ad hoc checks during an interactive session
Since validations accept arguments, the same validation can be attached to the same resource multiple times with different arguments. For example, local/run-tests might be attached to local/api-repo through project local/api-service with --coverage-threshold 90 and also through project local/staging with --coverage-threshold 70.
Attachment Resolution
When a plan executes, the system collects all applicable validations for each resource the plan accesses. The resolution algorithm:
- Collect direct validations: For every resource the plan accesses, collect all validations directly attached to that resource (no scope flag).
- Collect project-scoped validations: For every resource the plan accesses, collect all validations attached to that resource with a
--projectscope matching the plan's target project. - Collect plan-scoped validations: For every resource the plan accesses, collect all validations attached to that resource with a
--planscope matching the plan itself. - Union: Take the union of all collected validations per resource. If the same validation name appears from multiple scopes for the same resource (e.g., attached both directly and through a project), each attachment is treated as a distinct invocation — they may have different arguments and thus produce different results. When the same validation name appears from multiple scopes with identical arguments, the most specific scope wins (plan > project > direct) and the validation runs once.
- Child plan inheritance: Child plans (spawned via
subplan_spawnorsubplan_parallel_spawn) inherit their parent plan's project-scoped and plan-scoped resource/validation associations by default. Direct validations are collected independently based on which resources the child plan accesses (which may differ from the parent).
Attachment Identity
Each attachment is identified by a system-assigned ULID. This ULID is:
- Returned by
agents validation attachwhen the attachment is created. - Displayed by
agents tool show <VALIDATION>in the "Attached To" section and byagents project showin the validations list. - Required by
agents validation detach <ATTACHMENT_ID>to remove a specific attachment.
The ULID is necessary because the same validation can be attached to the same resource multiple times — directly, through different projects, through different plans, and even to the same resource and scope with different arguments. The ULID uniquely identifies which specific attachment to remove.
Attachment Arguments
When attaching a validation, optional arguments can be provided after the validation name:
$ agents validation attach --project local/api-service local/api-repo local/run-tests --coverage-threshold 90
$ agents validation attach --project local/staging-api local/api-repo local/run-tests --coverage-threshold 70
These arguments are stored with the attachment and passed to the validation tool's input_schema at execution time. This enables a single validation definition to be reused across scopes with different thresholds or configurations. If no arguments are provided, the validation uses its input_schema defaults.
Validation Lifecycle in Plan Execution
Validation participates in the plan lifecycle at well-defined points:
During Strategize
Validations are not executed during Strategize. However, the strategy actor is aware of which validations are attached (via the attachment resolution described above) and factors them into the strategy:
- The strategy may allocate time/tokens for validation fix-up loops.
- The strategy may note that certain validations are informational and thus non-blocking.
- The decision tree may include
validation_checkpointdecisions that mark points where validation should be run.
During Execute
Validation is the final step of the Execute phase. The execution actor's workflow proceeds as follows:
- Perform work: The actor executes the strategy — writing code, modifying resources, running tools — all within the sandbox.
- Collect applicable validations: The system resolves all validations from the resource-direct, project, and plan scopes (as described in Attachment Resolution above).
- Execute validations: Each applicable validation is invoked as a standard tool call. The validation reads the sandbox state and returns its structured result. Validations may be run in parallel since they are read-only and cannot interfere with each other.
- Process results:
- All required validations pass: Execution is complete. The plan transitions to the review/Apply phase.
- Any required validation fails: The execution actor enters the fix-then-revalidate loop (see Validation Failure Handling below).
- Informational validations fail: Results are recorded in the plan's validation summary. No fix attempts are made.
Validation is Not Run During Apply
Validation runs during Execute, not during Apply. By the time a plan reaches Apply, all required validations have already passed. Apply commits the sandbox changes to the real resources. If the user wants post-apply verification (e.g., integration tests against the real system), that should be implemented as a separate plan or an external CI pipeline — not as a validation attachment.
The rationale: Apply is a controlled commit step. Introducing validation during Apply would create a situation where committed changes might need to be rolled back, defeating the purpose of the sandbox model. The Execute phase is where all verification occurs, and Apply is the point of no return.
Validation Failure Handling
Because validations are tools, their execution follows the standard tool invocation flow. Each validation returns a structured JSON result with { "passed": true/false, "message": "...", "data": {...} }. The execution actor interprets these results based on the validation's mode.
Required Validation Failure
When a required validation returns "passed": false during Execute:
-
Diagnosis: The execution actor examines the validation's
messageanddatafields to understand the nature of the failure. A well-structureddatafield (e.g., with specific file paths, line numbers, error messages) enables more targeted fix attempts. -
Self-fix attempt: The execution actor attempts to fix the issue within the bounds of the current strategy. For example:
- A failing test: the actor reads the test output, identifies the broken assertion, and fixes the code.
- A lint error: the actor reads the lint report and corrects the formatting or style violation.
- A type error: the actor reads the type checker output and fixes the type mismatch.
-
Re-validation: After each fix attempt, the actor re-invokes the failing validation. If it passes, the loop ends. If it still fails, the actor examines the new
messageanddatafor the updated failure state. -
Retry limit: The fix-then-revalidate loop runs up to the configured retry limit (default: 3 attempts, configurable per plan or in the automation profile). After the limit is reached, self-fix stops.
-
Strategy revision: If the actor determines that the failure cannot be resolved within the current strategy's constraints (e.g., the strategy says "modify only
handler.py" but the test failure requires changes tomodel.py), it may request a strategy revision. This triggers a re-run of the Strategize phase for the affected subtree of the decision tree. Whether this happens automatically or requires user approval depends on the automation profile'sauto_strategy_revisionflag. -
Escalation: If strategy revision also fails, or if the automation profile requires human approval for strategy changes, the plan pauses and requests user guidance via
agents plan prompt:agents plan prompt <plan_id> "Try using mock objects for the database tests" -
Terminal failure: If the user does not intervene (or explicitly cancels), the plan fails with
state: failed. The sandbox is preserved for inspection.
The flow can be summarized as:
validate → fail → fix → re-validate → fail → fix → re-validate → ... → retry limit
→ request strategy revision → re-strategize → re-execute → validate
→ still failing → escalate to user → user provides guidance → resume
→ still failing → plan fails
Informational Validation Failure
When an informational validation returns "passed": false:
- The result (including
messageanddata) is recorded in the plan'svalidation_summary. - Execution continues normally — no fix attempts, no blocking, no escalation.
- The informational failure is visible in
agents plan show, plan summaries, and any rendering format. - Informational failures may trigger notifications (if configured) but never block plan progression.
Validation Error vs. Validation Failure
An important distinction:
- Validation failure (
passed: false): The validation ran successfully and determined the condition was not met. This is normal and handled by the mode-specific logic above. - Validation error: The validation itself failed to execute — a runtime exception, timeout, malformed return, or infrastructure failure. Validation errors are treated as required validation failures regardless of the validation's mode, because a validation that cannot run provides no assurance.
When a validation error occurs:
- The result is recorded with
passed: false, a system-generatedmessagedescribing the error, and the raw error details indata. - If the validation is required, the fix-then-revalidate loop attempts to resolve the error (e.g., by fixing a broken test configuration).
- If the validation is informational, the error is still recorded but does not block.
- Repeated validation errors (e.g., the validation tool itself is misconfigured) are surfaced prominently in the plan summary.
Validation in Actor Graphs
Because Validations are Tools, they can appear as tool nodes in actor graphs. This enables explicit, deterministic validation steps within workflows:
# Actor graph with explicit validation nodes
graph:
nodes:
- name: implement
type: agent
actor: local/code-writer
- name: run_tests
type: tool
tool: local/run-tests # This is a Validation (subtype of Tool)
- name: lint_check
type: tool
tool: local/lint-check # Also a Validation
- name: fix_issues
type: agent
actor: local/code-fixer
edges:
- [implement, run_tests]
- [implement, lint_check]
- condition: "not run_tests.passed or not lint_check.passed"
from: [run_tests, lint_check]
to: fix_issues
- [fix_issues, run_tests] # Retry loop
- [fix_issues, lint_check]
When a Validation appears as a tool node, its return value is available to downstream nodes via the standard graph data flow. The passed, message, and data fields can be used in edge conditions, passed as input to subsequent nodes, or logged for the plan summary.
This graph-based approach complements the attachment model: attached validations run automatically at the end of Execute (the system handles collection and invocation), while graph-embedded validations run at explicitly defined points in the workflow (the actor author controls placement). Both approaches can coexist — a plan may have attached validations that run at the end plus graph-embedded validations that run at intermediate checkpoints.
Validation and Skills
Validations can be included in skills just like any other tool. Since a Validation is a Tool, a skill YAML can reference it by name:
# Skill: local/python-quality
skill:
name: local/python-quality
description: "Python code quality tools and validations"
tools:
- local/run-tests # Validation (required)
- local/lint-check # Validation (required)
- local/type-check # Validation (required)
- local/format-code # Plain Tool (writes)
- local/run-benchmarks # Validation (informational)
When an actor references this skill, the Validations are available as tools that the LLM agent can call. This is distinct from validation attachment — being included in a skill makes the validation callable by the actor's LLM, but does not make it a required gate. Attachment determines gating; skill inclusion determines availability.
A common pattern: a team's skill includes their validations for on-demand use by the LLM (the actor can run tests proactively during development), while the same validations are also attached to the project (ensuring they run as a mandatory gate at the end of Execute even if the LLM forgot to run them).
Tool Wrapping
A Validation can wrap an existing Tool, reusing its implementation without duplicating code. This is the recommended approach when a plain Tool already performs the work a Validation needs (e.g., running tests, generating a report, scanning for vulnerabilities) and the Validation only needs to interpret the Tool's output as pass/fail.
The Problem
Consider a team that already has a registered Tool called local/run-tests — it runs the test suite and returns a structured report (test counts, coverage data, output logs). Now they want a Validation that gates execution on all tests passing. Without wrapping, they would need to duplicate the entire test-running implementation inside a new Validation YAML, differing only in the final pass/fail interpretation. This duplication creates maintenance burden: if the test runner configuration changes, both the Tool and the Validation must be updated.
The Solution: wraps + transform
A Validation YAML can declare a wraps field that references an existing registered Tool by name. At execution time, the system invokes the wrapped Tool, captures its output, and passes that output through a transform function that produces the Validation's structured return format ({ "passed", "message", "data" }).
# validations/tests-pass.yaml # Wraps the existing local/run-tests tool — no test logic duplicatedname: local/tests-pass description: "Validate that all unit tests pass (wraps local/run-tests)"
wraps: local/run-tests
transform: | def transform(tool_output): passed = tool_output.get("returncode") == 0 return { "passed": passed, "message": "All tests passed" if passed else f"Tests failed (exit {tool_output.get('returncode')})", "data": tool_output }
validation: mode: required
timeout: 600
When local/tests-pass is invoked (either by attachment or direct call):
- The system resolves
wraps: local/run-teststo the registered Tool. - The Validation's input arguments are mapped to the wrapped Tool's input arguments using the
argument_mapping(see below). If noargument_mappingis defined, the Validation's input arguments are passed through to the wrapped Tool as-is. - The wrapped Tool is invoked with the mapped arguments.
- The wrapped Tool's output is captured (whatever it returns — JSON, text, etc.).
- The
transformfunction is called with the captured output. - The
transformfunction returns the Validation's structured result ({ "passed", "message", "data" }).
This is opaque to the consumer — the caller sees only a standard Validation with its structured return format. The wrapping is an implementation detail.
What Gets Inherited
When a Validation uses wraps, the following properties are inherited from the wrapped Tool unless explicitly overridden in the Validation YAML:
| Property | Inherited? | Override behavior |
|---|---|---|
input_schema |
Yes | If the Validation YAML defines input_schema, it replaces the wrapped Tool's schema entirely. If omitted, the Validation accepts the same inputs as the wrapped Tool. When a custom input_schema is defined, an argument_mapping should also be provided to map the Validation's inputs to the wrapped Tool's expected inputs. |
argument_mapping |
No | Not inherited. When present, maps the Validation's input arguments to the wrapped Tool's input arguments. When absent, input arguments are passed through unchanged. See the Argument Mapping section. |
resource_slots |
Yes | If the Validation YAML defines resource_slots, they replace the wrapped Tool's slots. If omitted, the Validation inherits the same resource bindings. |
timeout |
Yes | If the Validation YAML specifies timeout, it overrides. If omitted, the wrapped Tool's timeout is used. |
description |
No | The Validation must provide its own description (it serves a different purpose than the wrapped Tool). |
source |
No | Ignored — the source is implicitly wrapped. The Validation does not declare source or code. |
capability |
No | Overridden — writes and checkpointable are forced to false regardless of the wrapped Tool's values. |
The transform Function
The transform field contains a Python function that converts the wrapped Tool's output to the Validation's return format. The function signature is:
def transform(tool_output) -> dict:
"""
Args:
tool_output: The wrapped Tool's return value. The type depends on the
Tool's implementation — typically a dict (for JSON-returning
tools) but may be a string or other type.
Returns:
A dict with at minimum {"passed": bool}. May also include
"message" (str) and "data" (any).
"""
The transform function runs in a sandboxed Python environment with no write access to the filesystem or network. It is a pure data transformation — it receives the Tool's output and produces the Validation's result. If the transform function raises an exception, the Validation is treated as an error (see Validation Error vs. Validation Failure).
If transform is omitted, the system applies a default transform that expects the wrapped Tool's output to already contain a passed field:
- If the output is a dict with a
passedboolean, it is used as-is (pass-through). - If the output is a dict without
passed, the validation errors with a message indicating the wrapped tool's output is not in validation format and atransformfunction is required. - If the output is not a dict, the validation errors similarly.
Argument Mapping (argument_mapping)
When a Validation wraps a Tool, the Validation may accept different input arguments than the wrapped Tool. The argument_mapping field specifies how the Validation's input arguments map to the wrapped Tool's expected input arguments. This is necessary when:
- The Validation's
input_schemadefines different field names than the wrapped Tool'sinput_schema. - The Validation accepts a subset of the wrapped Tool's arguments and wants to provide fixed values for the rest.
- The Validation renames arguments for clarity in the validation context.
The argument_mapping is a dictionary where keys are the wrapped Tool's input parameter names and values are either:
- A string referencing a Validation input parameter name (forwarded as-is).
- A fixed literal value (string, number, boolean) that is always passed to the wrapped Tool regardless of the Validation's inputs.
# Example: Validation wraps local/run-tests but renames arguments name: local/tests-pass wraps: local/run-testsargument_mapping: test_directory: source_dir # Forward Validation's "source_dir" → Tool's "test_directory" coverage_enabled: true # Always pass true for coverage_enabled verbose: false # Always pass false for verbose
transform: | def transform(tool_output): return {"passed": tool_output.get("returncode") == 0, "message": "Tests completed"}
validation: mode: required
When argument_mapping is omitted, the Validation's input arguments are passed through to the wrapped Tool unchanged. This is the common case when the Validation inherits the wrapped Tool's input_schema without modification (i.e., the Validation YAML does not define its own input_schema).
When argument_mapping is present, only the mapped arguments are forwarded to the wrapped Tool. Any Validation input arguments not referenced in the mapping are not passed to the wrapped Tool. Any wrapped Tool arguments not listed as keys in the mapping receive no value (and must either be optional in the Tool's schema or have defaults).
Read-Only Semantics for Wrapped Tools
Validations are always read-only, but the wrapped Tool may not be. The wrapping enforces read-only semantics:
- The wrapped Tool is invoked within the Validation's read-only execution context. Resource bindings are resolved with
access: read_onlyregardless of what the wrapped Tool'sresource_slotsspecify. - If the wrapped Tool's implementation attempts to write (e.g., mutate files, insert database rows), the sandbox enforces the read-only constraint and the invocation fails as a validation error.
- At registration time, if the wrapped Tool has
writes: true, the system emits a warning (not an error) advising the user that the wrapped Tool is marked as writing and may fail at runtime under read-only enforcement. This allows wrapping tools that are conservatively markedwrites: truebut whose actual behavior for the given inputs is read-only.
This design means wrapping is always safe from the Validation perspective — the worst case is a runtime error, never an unintended write.
Multiple Validations Wrapping the Same Tool
The same Tool can be wrapped by multiple Validations with different transform functions and different modes. This is a common pattern:
# The base tool runs tests and produces a comprehensive report # Tool: local/run-tests (already registered)
# Validation 1: All tests must pass (required) # validations/tests-pass.yaml name: local/tests-pass description: "All unit tests must pass" wraps: local/run-tests transform: | def transform(tool_output): return { "passed": tool_output.get("returncode") == 0, "message": f"{tool_output.get('tests_passed', 0)}/{tool_output.get('tests_run', 0)} tests passed", "data": tool_output } validation: mode: required
# Validation 2: Coverage must exceed threshold (informational for now)
# validations/coverage-check.yaml
name: local/coverage-check
description: "Coverage must exceed threshold (advisory)"
wraps: local/run-tests
transform: |
def transform(tool_output):
coverage = tool_output.get("coverage_percent", 0)
threshold = 80
return {
"passed": coverage >= threshold,
"message": f"Coverage: {coverage}% (threshold: {threshold}%)",
"data": {"coverage_percent": coverage, "threshold": threshold}
}
validation:
mode: informational
Both validations wrap the same local/run-tests tool but extract different signals. When both are attached to the same project, each Validation independently invokes the wrapped Tool and applies its own transform function. A future optimization may deduplicate these invocations when the forwarded arguments are identical (see Validation Ordering and Parallelism).
When to Wrap vs. When to Write From Scratch
| Scenario | Approach |
|---|---|
| An existing Tool already does the work; you just need pass/fail | Wrap it — use wraps + transform |
| Multiple Validations need different interpretations of the same Tool's output | Wrap it multiple times with different transforms |
| The validation logic is simple and self-contained (e.g., run a shell command, check exit code) | Write from scratch — use source: custom with inline code |
| The validation needs logic that the existing Tool doesn't provide (different inputs, different execution) | Write from scratch — the Tool's implementation isn't relevant |
| The existing Tool has side effects that are incompatible with read-only enforcement | Write from scratch — wrapping would fail at runtime |
Validation Data Model
The validation-related fields stored in the plan data model:
| Field | Location | Description |
|---|---|---|
validation_summary |
plan.execution |
Array of validation results collected during Execute. Each entry includes the validation name, mode, passed, message, data, execution duration, and attempt number. |
final_validation_results |
plan.apply |
Snapshot of the final validation state at the time execution completed (all required validations passing). Identical to the last state of validation_summary but stored separately for quick reference. |
validation_attempts |
plan.execution |
Total number of validation invocations across all fix-then-revalidate loops. Useful for understanding how much effort was spent on validation remediation. |
validation_fix_history |
plan.execution |
Per-validation log of fix attempts. Each entry records: the validation that failed, the fix the actor applied, and whether the subsequent re-validation passed. Enables auditing of the fix-then-revalidate process. |
A validation result entry in validation_summary:
{
"validation": "local/run-tests",
"mode": "required",
"passed": true,
"message": "All 247 tests passed, 94% coverage",
"data": {
"tests_run": 247,
"tests_passed": 247,
"coverage_percent": 94.2
},
"duration_ms": 12400,
"attempt": 2,
"attachment_id": "01HXM5A1B2C3D4E5F6G7H8J9K0",
"attachment_resource": "local/api-repo",
"attachment_scope": "project",
"attachment_scope_target": "local/api-service"
}
Validation Ordering and Parallelism
Since validations are read-only, they are inherently safe to run in parallel. The execution runtime may parallelize validation invocations subject to:
- Resource constraints: Validations that require significant compute (e.g., full test suites) may be serialized to avoid resource exhaustion.
- Timeout budgets: Each validation has its own
timeout(from the tool configuration). The total validation phase has an implicit budget derived from the sum of individual timeouts, but the runtime may use parallelism to finish faster. - Dependency hints: A future extension may allow declaring ordering dependencies between validations (e.g., "run lint before tests" to fail fast on formatting issues). Currently, all validations run independently.
Wrapped Tool deduplication (deferred): When multiple Validations wrap the same Tool (via wraps), a future optimization could invoke the wrapped Tool once and fan out its output to each Validation's transform function. However, deduplication is only valid when multiple Validations wrap the same Tool and receive the exact same forwarded arguments (after argument_mapping resolution) and are evaluated in the same calling chain. This complexity — particularly around argument equivalence and cache invalidation during fix-then-revalidate loops — means deduplication is deferred for future design. The current implementation invokes the wrapped Tool separately for each Validation that wraps it.
The recommended execution strategy is:
- Run all validations in parallel (each Validation independently invokes its wrapped Tool if applicable).
- Collect results.
- If any required validation fails, fix and re-run only the failing validations (not all of them).
- Repeat until all pass or retry limit is reached.
Registering and Managing Validations
Validations are registered via agents validation add and attached via agents validation attach. These are the only validation-specific CLI commands. All other management operations use the standard agents tool commands:
| Operation | Command | Notes |
|---|---|---|
| Register | agents validation add --config <FILE> |
Validation-specific. Config file defines all properties including mode (required/informational). |
| Update | agents validation add --config <FILE> --update |
Uses the --update flag to overwrite existing registration. |
| Attach | agents validation attach [--project|--plan] <RESOURCE> <VALIDATION> [args...] |
Returns an attachment ULID. Resource is always required; project/plan scope is optional. |
| Detach | agents validation detach [--yes] <ATTACHMENT_ID> |
Uses the attachment ULID, not the validation name. |
| List | agents tool list --type validation |
Shared with tools. Use --type validation to filter. |
| Show | agents tool show <NAME> |
Shows validation-specific fields when the entry is a Validation. |
| Remove | agents tool remove <NAME> |
Shared with tools. Automatically detaches from all scopes. |
See the agents validation and agents tool CLI Reference sections for full command details and examples.
Typical Validation Workflow
A typical workflow for setting up validations on a project:
# 1. Define validation YAML files # (see Configuration > Validation Configuration Files for schema and examples)# 2. Register validations agents validation add --config ./validations/run-tests.yaml agents validation add --config ./validations/lint-check.yaml agents validation add --config ./validations/type-check.yaml agents validation add --config ./validations/check-bundle-size.yaml
# 3. Attach to resource through a project (active only when this resource is accessed through this project) agents validation attach --project local/api-service local/api-repo local/run-tests agents validation attach --project local/api-service local/api-repo local/type-check agents validation attach --project local/api-service local/api-repo local/check-bundle-size
# 4. Attach directly to a resource (always active for any plan accessing this resource) agents validation attach local/api-repo local/lint-check
# 5. Verify setup agents tool list --type validation --namespace local agents tool show local/run-tests
# 6. Run a plan — validations execute automatically at end of Execute phase agents plan use local/implement-feature local/api-service
Design Rationale
Why Validation extends Tool rather than being a separate concept:
- Reuses the entire tool infrastructure (registry, bindings, lifecycle, sources, schemas) without duplication.
- Validations can be composed into skills, used in actor graphs, and invoked by LLM agents — all for free.
- A single namespace and registry avoids the complexity of managing two parallel systems.
- Tool-aware features (resource bindings, MCP integration, Agent Skills) apply to validations automatically.
Why read-only is enforced rather than advisory:
- Validations that modify state defeat the purpose of the sandbox model.
- Read-only guarantees enable safe parallelization and retry without side effects.
- It makes validations categorically safe — no risk analysis needed for running a validation.
Why attachments use ULIDs rather than name+scope pairs:
- A validation may be attached to the same scope multiple times with different arguments.
- ULIDs provide an unambiguous handle for removal without complex composite keys.
- Consistent with the ULID-based identity model used for plans, resources, and decisions.
Why validation runs during Execute, not Apply:
- Apply is a controlled commit step. Validation failures during Apply would require rolling back committed changes.
- The sandbox model ensures all verification occurs before the point of no return.
- Post-apply verification (integration tests, smoke tests) is a separate concern best handled by external CI or follow-up plans.
Skills
!!! adr "Architecture Decision" The skill composition model, skill inclusion, and tool bundling are defined in ADR-012: Skill System.
What a Skill Is
A skill is a namespaced, reusable collection of tools that is registered in the system via its own YAML configuration file and managed through agents skill CLI commands. Skills are the unit of capability composition in CleverAgents — they define what an actor can do by assembling tools into coherent, reusable bundles.
A skill is not a single tool. It is a container that references one or more tools (by name from the Tool Registry) and/or defines anonymous inline tools, along with metadata describing the collection's purpose, capabilities, and safety characteristics. Tools are independently registered, callable operations (see the Tools section above). Skills organize them into reusable groups.
Key properties of a skill:
- Named and namespaced: Skills follow the same
<namespace>/<name>naming convention as actors, tools, actions, and plans (e.g.,local/github-ops,cleverthis/file-management,local/deploy-tools). - Defined in their own YAML files: Skills are NOT defined inline in actor configurations. They have their own configuration files and are managed as independent, reusable entities.
- A collection of tools: Each skill references named tools from the Tool Registry and/or defines anonymous inline tools. Skills can also expose tools from MCP servers, Agent Skills Standard folders, and built-in tool groups.
- Hierarchically composable: A skill can include other skills by reference, inheriting all of their tools. This enables layered composition — a "full-stack" skill might include a "file-ops" skill, a "git-ops" skill, and a "github" skill. When including a sub-skill, individual tool metadata can optionally be overridden.
- Referenced by actors: Actors reference skills by fully-qualified name. The actor's graph gains access to all tools within the referenced skills.
The Skill / Tool Distinction
@startuml
skinparam packageStyle rectangle
skinparam defaultFontSize 12
skinparam componentFontSize 12
package "Skill: local/devops-toolkit" as Skill {
package "builtin: file_operations" as FileOps {
component [read_file()] as T1
component [write_file()] as T2
component [edit_file()] as T3
}
package "builtin: git_operations" as GitOps {
component [git_status()] as T4
component [git_diff()] as T5
}
package "mcp: github-server" as GH {
component [create_issue()] as T6
component [create_pr()] as T7
component [list_repos()] as T8
}
package "custom" as Custom {
component [run_migrations()] as T9
}
package "Included Skills" as Includes {
component [local/pdf-processing\n(adds pdf tools)] as I1
component [local/data-analysis\n(adds analysis tools)] as I2
}
}
@enduml
Tools are independently registered, atomic units of execution (see the Tools section above for full details). Each tool has:
- A namespaced name, description, and JSON Schema for inputs/outputs
- Its own YAML configuration file, registered via
agents tool add - A source type (mcp, agent_skill, builtin, custom)
- Capability metadata (read_only, writes, checkpointable, etc.)
- A lifecycle:
discover(),activate(),execute(params, ctx),deactivate()
Skills are the organizational units. Each skill has:
- A namespaced name
- A YAML configuration file defining its tool composition
- Zero or more named tool references (pointing to independently registered tools in the Tool Registry)
- Zero or more anonymous inline tools (one-off tools defined directly in the skill YAML)
- Zero or more included child skills (whose tools are merged in, with optional per-tool metadata overrides)
- Tool sources: MCP servers, Agent Skills folders, built-in tool groups
- Metadata (description, capability summary)
When an actor references a skill, it gains access to the flattened set of all tools — named tool references, anonymous tools, tools from MCP/Agent Skills/builtins, and those inherited from included child skills.
Skill Configuration (YAML)
Skills are defined in their own YAML configuration files, separate from tool and actor configurations. A skill YAML file declares which registered tools it includes (by name), which other skills it includes, and optionally defines anonymous inline tools:
# File: skills/devops-toolkit.yaml cleveragents: version: "3.0"skill: name: local/devops-toolkit description: "Full-stack development tools for file ops, git, GitHub, and deployment"
# ── Named Tool References ─────────────────────────────── # Reference independently registered tools by name. # These tools must already be registered via
agents tool add. # Optional metadata overrides can be applied per tool. tools: - local/run-migrations # Simple reference, use as registered - name: local/deploy-staging # Reference with metadata override override: capability: human_approval_required: true # Require approval in this skill context# ── Include other skills ───────────────────────────────── # All tools from included skills become part of this skill. # Included skills must already be registered in the system. # Individual tools from included skills can have metadata overridden. includes: - local/file-ops # built-in file + directory + search tools - local/git-ops # built-in git tools - name: local/github # MCP-based GitHub tools, with per-tool overrides tool_overrides: - tool: local/create-github-issue override: capability: write_scope: [github:issues:org-only]
# ── MCP Server Tools ───────────────────────────────────── # Connect to MCP servers and expose their tools. # Tools discovered from MCP servers are auto-registered in the # Tool Registry if not already present. mcp_servers: - name: linear command: "npx @anthropic/mcp-linear" env: LINEAR_API_KEY: "${LINEAR_API_KEY}" # Optional: override inferred capability metadata per tool overrides: - tool: create_issue writes: true write_scope: [linear:issues] - tool: list_issues read_only: true
# ── Agent Skills (SKILL.md folders) ────────────────────── # Each Agent Skill folder is loaded as a composite tool. # The agent discovers it via metadata, activates it by # loading SKILL.md instructions, and follows them. agent_skills: - path: ./skills/code-review-checklist sandbox_policy: none
# ── Built-in Tool Groups ───────────────────────────────── # Opt-in to built-in tool groups provided by CleverAgents. builtins: - group: shell_operations
# ── Anonymous Tools ────────────────────────────────────── # Inline tool definitions for one-off, skill-specific operations. # Same format as a named tool YAML body but without a name. # These are NOT registered in the Tool Registry and are NOT reusable. anonymous_tools: - description: "One-off cleanup for legacy migration artifacts" input_schema: type: object properties: directory: { type: string } capability: writes: true checkpointable: true checkpoint_scope: file code: | import os, glob directory = params["directory"] removed = [] for f in glob.glob(os.path.join(ctx.sandbox.root, directory, "*.legacy")): os.remove(f) removed.append(f) return {"removed": removed, "count": len(removed)}
Here is an example of a simpler skill that wraps only built-in tools, suitable for common reuse:
# File: skills/file-ops.yaml cleveragents: version: "3.0"skill: name: local/file-ops description: "File and directory operations"
builtins: - group: file_operations # read, write, edit, delete, move, copy - group: directory_operations # create, list, delete dirs
And an example of a skill that is purely MCP-based:
# File: skills/github.yaml cleveragents: version: "3.0"skill: name: local/github description: "GitHub operations via MCP"
mcp_servers: - name: github command: "npx @anthropic/mcp-github" env: GITHUB_TOKEN: "${GITHUB_TOKEN}" overrides: - tool: create_issue writes: true write_scope: [github:issues] checkpointable: false - tool: create_pull_request writes: true write_scope: [github:pulls] checkpointable: false - tool: list_repos read_only: true - tool: get_file_contents read_only: true
Skill Hierarchy and Composition
Skills can include other skills via the includes field. When a skill includes another, all tools from the child skill (and transitively, all tools from any skills it includes) become part of the parent skill's flattened tool set.
local/full-stack-dev ├── includes: local/file-ops │ └── builtins: file_operations, directory_operations ├── includes: local/git-ops │ └── builtins: git_operations ├── includes: local/github │ └── mcp_servers: github (create_issue, create_pr, list_repos, ...) ├── agent_skills: code-review-checklist └── tools: local/run-migrations, local/deploy-staging (named tool refs)
Flattened tool set available to actors referencing local/full-stack-dev: read_file, write_file, edit_file, delete_file, move_file, copy_file, create_directory, list_directory, delete_directory, git_status, git_diff, git_log, git_blame, create_issue, create_pr, list_repos, get_file_contents, code-review-checklist (agent skill), local/run-migrations, local/deploy-staging (named tools)
Rules for skill composition:
- Circular includes are forbidden. The system validates the include graph at registration time and rejects cycles.
- Tool name conflicts: If two included skills provide tools with the same name, the conflict is resolved by qualification — the tool must be referenced as
<skill_name>.<tool_name>(e.g.,local/github.create_issuevslocal/linear.create_issue). Direct tools (defined in the skill itself) take precedence over included tools. - Included skills must be registered before the including skill can be added. The
agents skill addcommand validates this. - Depth is unlimited but the flattened tool set is computed at registration time and cached. Deep hierarchies do not incur runtime overhead.
Skill Registration and Management
Skills are managed through the agents skill CLI commands. Named tools referenced by skills must first be registered via agents tool add (see the Tools section):
# First, register any named tools the skill will reference agents tool add --config ./tools/run-migrations.yaml agents tool add --config ./tools/deploy-staging.yaml# Then register the skill (which references those tools by name) agents skill add --config ./skills/devops-toolkit.yaml
# Update an existing skill (re-reads the config file, overwrites registration) agents skill add --config ./skills/devops-toolkit.yaml --update
# List all registered skills agents skill list
# Show details for a skill (tools, includes, metadata) agents skill show local/devops-toolkit
# List all tools provided by a skill (flattened, including from child skills) agents skill tools local/devops-toolkit
# Remove a skill agents skill remove local/devops-toolkit
Once registered, a skill is available to be referenced by any actor configuration. Skills persist in the database (local or server) and follow the same namespace rules as actors, tools, and actions.
Actor References to Skills and Tools
Actors reference skills by name to make collections of tools available for LLM tool-calling. Additionally, actor graphs can include tool nodes that directly reference named tools or define anonymous inline tools (see Nodes in the Graph in the Actor section).
The actor's configuration lists which skills it should have access to:
# File: actors/code-assistant.yaml cleveragents: version: "3.0" default_actor: code_assistantactors: code_assistant: type: llm config: actor: anthropic/claude-3-opus temperature: 0.3 system_prompt: | You are a code assistant with access to file, git, and GitHub tools. Current task: {{ context.task_description }}
# Reference skills by fully-qualified name. # All tools from these skills become available to this actor. skills: - local/file-ops - local/git-ops - local/github
# Or reference a single composite skill that includes all of the above full_stack_assistant: type: llm config: actor: anthropic/claude-3-opus system_prompt: | You are a full-stack development assistant.
<span style="color: cyan; font-weight: 600;">skills</span>: - local/full-stack-dev # includes file-ops, git-ops, github, etc.
Server-qualified skill references: When connected to multiple servers, skills can be disambiguated with a server prefix, same as actors:
skills:
- dev:freemo/custom-analysis # from dev server, personal namespace
- prod:cleverthis/deploy-tools # from prod server, org namespace
- local/file-ops # local skill
Skill Registry
To enable plan validation and discovery at scale, CleverAgents maintains two registries that work together:
-
Tool Registry — a persistent catalog of all independently registered tools (described in the Tools section above). Managed via
agents tool add/remove/list/show. -
Skill Registry — a persistent catalog of all registered skills and their flattened tool sets. The Skill Registry composes its tool sets by resolving named tool references from the Tool Registry, incorporating anonymous inline tools, and merging tools from included child skills.
@startuml
skinparam classAttributeIconSize 0
skinparam classFontSize 13
skinparam defaultFontSize 12
class SkillRegistry {
- skillIndex : Map<String, SkillRecord>
--
+ add(config_path) : SkillRecord
+ update(name, config_path) : SkillRecord
+ remove(name) : void
+ lookup(name) : SkillRecord
+ list(filters) : SkillRecord[]
+ tools(name) : ToolDescriptor[]
+ validate_plan(plan) : ValidationResult
+ refresh(name) : void
}
class SkillRecord {
+ name : String
+ description : String
+ config_path : String
+ includes : List<String>
+ tool_refs : List<String>
+ anonymous_tools : List<ToolDef>
+ flattened_tools : List<ToolDescriptor>
+ overrides : Map<String, Object>
+ capability_summary : CapabilitySummary
}
SkillRegistry "1" *-- "0..*" SkillRecord : indexes >
note right of SkillRegistry
**Populated by:**
- agents skill add CLI command
- Dynamic refresh on MCP notifications
**Depends on:**
- Tool Registry (resolve named tool references)
**Consumed by:**
- Actor activation
- Plan validation
- Agent context injection
end note
@enduml
Both registries persist in the database (local SQLite or server). MCP server tools are refreshed dynamically when notifications/tools/list_changed events are received.
Session
!!! adr "Architecture Decision" Session persistence, session lifecycle, and session-plan relationships are defined in ADR-020: Session Model.
What a Session Is
A session is a user's interactive thread with CleverAgents across time.
A session should:
- maintain conversational continuity,
- store plan references,
- persist memory (if enabled),
- provide a UI anchor (CLI invocation, TUI workspace, web session).
Session and Memory Persistence
The notes include a known issue: conversation history can be lost between CLI invocations depending on connection string configuration, implying the system needs a stable memory service backend.
Therefore, CleverAgents should specify:
- sessions have stable IDs
- sessions can be resumed
- session storage backend is configured explicitly
- if session persistence is disabled, the UX should be explicit about it (no silent history loss)
Server
!!! adr "Architecture Decision" The server architecture, multi-user support, and local/server mode duality are defined in ADR-023: Server Mode.
What a Server Is
A server is an optional mode that enables:
- multi-user access
- shared org namespaces (
<username>/and<orgname>/) - persistent plan records
- remote plan execution
- governance
Single-user local mode is the default (so setup is easy), but the architecture anticipates server mode for shared skills and org-level actions.
Client-Only vs Server Mode
| Mode | Description | Plan Execution |
|---|---|---|
| Client-only | No server connection. All data in local database. | Always local |
| Server mode | Connected to a CleverAgents server. Namespaced items sync. | Local or server |
It is possible to run a client with no server at all. Server is optional.
Plan Execution Location
Where a plan executes depends on the project type:
| Project Type | Client Execution | Server Execution |
|---|---|---|
| Local (has local-only resources) | ✅ Yes | ❌ No (server can't access local resources) |
| Remote (all resources remotely accessible) | ✅ Yes | ✅ Yes |
When acting on local projects, the client must be running because only the client can access local resources.
Remote projects can execute on either:
- The client (if user prefers local execution)
- The server (for long-running plans, since client may be transient)
Server Execution Benefits
Server execution is useful when:
- Plans take a long time to execute
- Client may disconnect (laptop closes, network issues)
- Multiple team members need to monitor plan progress
- Centralized logging and auditing required
No Plan Queuing
Plans are not queued. When a plan is used on projects and executed, it runs immediately. There is no worker queue or delayed execution model.
Multi-user Risks and Prompt Injection
Prompt injection isn't critical in single-user mode but becomes important for multi-user server environments.
Server mode must include:
- access boundaries
- prompt sanitization / safe templating
- resource access controls
- auditing
Resources
!!! adr "Architecture Decision" The resource model, resource types, DAG structure, and resource lifecycle are defined in ADR-008: Resource System.
A resource is an independently registered entity representing anything that a plan can reason about or manipulate — git repositories, filesystems, databases, APIs, documents, and more. Resources are first-class citizens in CleverAgents, managed through the agents resource CLI commands and stored in the Resource Registry.
CleverAgents extends the MCP resource concept to support both read AND write operations (MCP resources are read-only). Unlike MCP's flat resource model, CleverAgents resources form a directed acyclic graph (DAG) with parent/child relationships, support physical vs virtual distinction for content identity tracking, and are governed by a resource type system that constrains their structure and behavior.
Resources are registered independently of projects. Projects link to resources from the Resource Registry — a resource can be linked to multiple projects, enabling shared resources across teams and workflows.
What a Resource Is
A resource has:
- Name and namespace: User-added resources follow the same
<namespace>/<name>naming convention as actors, tools, skills, and other entities (e.g.,local/api-repo,cleverthis/staging-db). Auto-discovered child resources do not have names — they are identified by ULID only. Every resource (whether user-added or auto-discovered) always has a system-assigned ULID. - Resource type: Every resource has a type (e.g.,
git,fs-mount,git-branch,fs-file) that determines its properties, CLI arguments, allowed parent/child relationships, sandbox strategy, and handler implementation. - Physical or virtual nature: Resources are either physical (a specific, concrete manifestation) or virtual (an abstract identity linking equivalent physical resources). This is determined by the resource type.
- Value/properties: Type-specific properties (a file path, a URL, a connection string, a commit hash, etc.).
- Parent/child relationships: Resources form a DAG. A resource can have multiple parents and multiple children, subject to type constraints.
- Capabilities: Whether the resource is readable, writable, sandboxable, and checkpointable.
Resource Types
A resource type is a schema-level definition that constrains a category of resources. Resource types define:
- CLI arguments: What arguments
agents resource add <type>accepts (for user-addable types), including which are required vs optional and validation rules. - Physical or virtual: Whether instances of this type are physical or virtual resources.
- Allowed parent types: What resource types are valid parents.
- Allowed child types: What resource types are valid children, and whether they are auto-discovered or manually linkable.
- Auto-discovery behavior: What child resources are automatically created when an instance of this type is registered (e.g., registering a
git-checkoutresource auto-discovers agitchild and anfs-directorychild for the worktree root; thegitchild in turn auto-discovers remotes, branches, commits, and tree entries). - User addable: Whether users can create instances of this type directly via
agents resource add <type>. Types withuser_addable: falseare only generated as auto-discovered children of other resources. - Sandbox strategy: The default sandboxing approach for instances of this type.
- Handler: The resource handler implementation that provides read/write/sandbox/checkpoint operations.
Built-in Resource Types
Built-in types are organized into three layers: git (version control structure), git-checkout (a composition that bridges git metadata with a local directory), and filesystem (physical files on disk). Virtual types link equivalent physical resources across these layers through content/identity matching. There are 24 built-in types total (15 physical + 9 virtual), of which 4 are user-addable (git-checkout, git, fs-mount, fs-directory).
Each table below includes the full parent/child relationship constraints. Allowed Parents lists what types may be a parent of this type, with cardinality (how many parents of that type are allowed) and whether the relationship is required or optional. Allowed Children lists what types may be children, with cardinality and whether they are required or optional. 0..* means zero or more, 1 means exactly one (required), 0..1 means zero or one (optional).
Physical types — Git layer (version control structure — every instance is a concrete manifestation in a specific repository):
| Type | User Addable | Sandbox | Allowed Parents | Allowed Children | Description |
|---|---|---|---|---|---|
git |
yes | none |
git-checkout (0..1, optional) |
git-remote (0.., optional), git-branch (0.., optional), git-tag (0.., optional), git-commit (0.., optional), git-stash (0.., optional), git-submodule (0.., optional) |
A git repository — the object database, refs, and full history. Accessible via a local .git directory path or a remote URL. Every git resource is a specific, concrete repo instance (local or hosted on a remote server). |
git-remote |
no | none |
git (1, required) |
(none) | A remote URL configured on a git repo (e.g., origin → https://github.com/org/repo). Auto-discovered child of git. Optionally linked as a child of a remote virtual resource when the same URL appears in multiple repos. |
git-branch |
no | (inherits) | git (1, required) |
git-commit (0..*, optional) |
A named branch ref (e.g., main, feature/auth). Auto-discovered child of git. Optionally linked as a child of a branch virtual resource when the same branch name + HEAD exists in multiple repos. |
git-tag |
no | (inherits) | git (1, required) |
(none) | A tag ref — lightweight or annotated (e.g., v1.0.0). Auto-discovered child of git. Optionally linked as a child of a tag virtual resource when the same tag name + target exists in multiple repos. |
git-commit |
no | (inherits) | git-branch (1..*, required), git (1, required) |
git-tree (1, required) |
A specific commit object. Auto-discovered child of git-branch (a commit can belong to multiple branches). Each commit contains exactly one root git-tree. Also a direct child of git for ancestry traversal. Optionally linked as a child of a commit virtual resource when the same commit hash exists in multiple repos (e.g., shared history between fork and upstream). |
git-tree |
no | (inherits) | git-commit (1..*, required) |
git-tree-entry (0.., optional), git-tree (0.., optional) |
A tree object — a directory listing at a specific commit. Each tree contains entries (blobs and subtrees). Auto-discovered child of git-commit. Trees are recursive: a tree can contain subtrees as children. Optionally linked as a child of a tree virtual resource when the same tree hash exists across commits/repos. |
git-tree-entry |
no | (inherits) | git-tree (1, required) |
(none) | A blob entry in a tree — a specific file's content at a specific path and mode. Auto-discovered child of git-tree. Represents the leaf of git's content-addressable storage. Optionally linked as a child of a file virtual resource when byte-identical content with the same name and permissions exists in the filesystem layer. |
git-stash |
no | (inherits) | git (1, required) |
(none) | A stash entry (e.g., stash@{0}). Auto-discovered child of git. Represents work-in-progress saved via git stash. |
git-submodule |
no | (inherits) | git (1, required) |
(none) | A submodule reference — a pointer to another git repository at a specific commit and path. Auto-discovered child of git. Optionally linked as a child of a submodule virtual resource when the same submodule URL + path exists in multiple repos. |
Physical types — Git checkout (composition: git metadata + local worktree directory):
| Type | User Addable | Sandbox | Allowed Parents | Allowed Children | Description |
|---|---|---|---|---|---|
git-checkout |
yes | git_worktree |
(none — always a top-level resource) | git (1, required), fs-directory (1, required) |
A locally checked-out git repository. The composition type — auto-discovers a git child (the repo's object database, branches, tags, commits, trees, and remotes) and an fs-directory child (the worktree root directory, e.g., /home/user/projects/my-app). Most users register this type. The worktree root IS a directory, not a mount point, so git-checkout composes with fs-directory directly. |
Physical types — Filesystem layer (files on disk, used by git checkouts and standalone directories alike):
| Type | User Addable | Sandbox | Allowed Parents | Allowed Children | Description |
|---|---|---|---|---|---|
fs-mount |
yes | copy_on_write |
(none — always a top-level resource) | fs-directory (1, required) |
A physical mount point on the local system (e.g., /mnt/data, /home). Represents the mount itself — the filesystem type (ext4, btrfs, etc.) is a property. The root directory is a required auto-discovered fs-directory child. Used for registering entire mount points or storage volumes. |
fs-directory |
yes | copy_on_write |
git-checkout (0..1, optional), fs-mount (0..1, optional), fs-directory (0..1, optional) |
fs-directory (0.., optional), fs-file (0.., optional), fs-symlink (0.., optional), fs-hardlink (0.., optional) |
A directory on the filesystem. Can be the root child of fs-mount (mount point root), the worktree root child of git-checkout, a subdirectory child of another fs-directory, or a standalone user-registered directory. Optionally linked as a child of a directory virtual resource when equivalent directory content exists elsewhere. When user-addable, accepts --path flag. |
fs-file |
no | (inherits) | fs-directory (1, required) |
(none) | A regular file on the local filesystem. Auto-discovered child of fs-directory. Optionally linked as a child of a file virtual resource when byte-identical content with the same name and permissions exists elsewhere (in another fs-file or a git-tree-entry). |
fs-symlink |
no | (inherits) | fs-directory (1, required) |
(none) | A symbolic link on the local filesystem. Auto-discovered child of fs-directory. Optionally linked as a child of a symlink virtual resource when a symlink with the same name and target exists elsewhere. |
fs-hardlink |
no | (inherits) | fs-directory (1, required) |
(none) | A hard link on the local filesystem — a file with link count > 1. Auto-discovered child of fs-directory. The system tracks hard link relationships by inode to avoid treating the same underlying data as distinct resources. |
Virtual types (abstract identity types that link equivalent physical resources — never user-addable, no sandbox):
Virtual types use simple names that mirror their physical counterparts. A virtual resource answers the question: "where else does this same thing exist?" Two physical resources share a virtual parent when they are equivalent by the virtual type's criteria (same content, same name, same permissions, same hash, etc.).
| Type | Allowed Children | Description |
|---|---|---|
file |
fs-file (0.., optional), git-tree-entry (0.., optional) |
Cross-layer file identity. Links physical fs-file and git-tree-entry resources that represent the same file — identical content bytes, filename, and permissions. Answers: "this file in the working tree and this blob in git's tree are the same file." |
directory |
fs-directory (0.., optional), git-tree (0.., optional) |
Cross-layer directory identity. Links physical fs-directory and git-tree resources whose recursive contents are equivalent. Answers: "this directory on disk matches this tree object in git." |
symlink |
fs-symlink (0.., optional), git-tree-entry (0.., optional) |
Cross-layer symlink identity. Links physical fs-symlink and git-tree-entry (mode 120000) resources with the same name and target. |
commit |
git-commit (0..*, optional) |
Cross-repo commit identity. Links git-commit resources across different repos that share the same commit hash — typically a fork and its upstream. Answers: "this commit exists in both repos." |
branch |
git-branch (0..*, optional) |
Cross-repo branch identity. Links git-branch resources across different repos with the same branch name and HEAD commit hash. |
tag |
git-tag (0..*, optional) |
Cross-repo tag identity. Links git-tag resources across different repos with the same tag name and target object. |
remote |
git-remote (0..*, optional) |
Cross-repo remote identity. Links git-remote resources across different repos that point to the same URL. Answers: "these repos share the same upstream." |
submodule |
git-submodule (0..*, optional) |
Cross-repo submodule identity. Links git-submodule resources across different repos with the same submodule URL and path. |
tree |
git-tree (0..*, optional) |
Cross-repo/cross-commit tree identity. Links git-tree resources across different commits or repos that have the same tree hash — identical directory structure and contents. |
Key design notes:
- Virtual types are never user-addable — they are created and maintained automatically by the system when equivalent physical resources are detected.
- Virtual types have no sandbox strategy because they represent abstract identities, not concrete locations. Tools always operate on the physical children.
- Physical types list virtual parents in their Allowed Parents column (via the "optionally linked" descriptions). For example, an
fs-filecan be linked as a child of afilevirtual resource when its content matches agit-tree-entry. This is how the DAG connects physical and virtual layers. gitrepresents a specific git repository instance — whether hosted remotely (e.g., on GitHub's servers) or locally (a.gitdirectory on disk). It is physical because it is a concrete manifestation that exists somewhere, not an abstract identity. Agitresource can be created from a remote URL alone (accessing the object database via the git protocol) or from a local.gitdirectory.git-checkoutis the type most users register for repos they've cloned. It composes agitchild (repo metadata) and anfs-directorychild (the worktree root directory). The worktree root is a directory on an existing filesystem — NOT a separate mount point — sogit-checkoutlinks directly tofs-directory, notfs-mount. This means the worktree's files use the samefs-directoryandfs-filetypes regardless of how they were created.git-commithas agit-treechild (the root tree object), which in turn hasgit-tree-entryand nestedgit-treechildren. This properly models git's internal object structure: commits point to trees, trees contain entries (blobs) and subtrees.fs-mountrepresents a physical mount point, not a directory. The filesystem type (ext4, btrfs, etc.) is a property on the mount. The root directory is an auto-discoveredfs-directorychild. Usefs-mountfor registering mount points and storage volumes. Usefs-directory(user-addable) for registering arbitrary directories.fs-directoryis user-addable, allowing users to register standalone directories (e.g., a build output folder, a config directory) without wrapping them in anfs-mount. Anfs-directoryregistered standalone or as a child offs-mount/git-checkoutuses the same type and auto-discovers the samefs-file,fs-symlink,fs-hardlink, and nestedfs-directorychildren.git-tree-entryrepresents a blob in git's tree — a specific file's content at a specific path and mode. The corresponding file on disk (if checked out) is anfs-file. Thefilevirtual type bridges these two when their content, name, and permissions are identical.- 4 user-addable types:
git-checkout(local repo),git(remote or local metadata-only),fs-mount(mount point),fs-directory(arbitrary directory).
Built-in Type Hierarchy
The parent-child relationships between built-in resource types form three layers — git structure, git checkout (composition), and filesystem — bridged by virtual identity types:
@startuml
skinparam defaultFontSize 11
skinparam objectFontSize 11
skinparam packageStyle rectangle
package "GIT-CHECKOUT (composition)" as GC #LightBlue {
object "git-checkout" as gc {
/home/user/projects/myapp
}
}
package "GIT STRUCTURE" as GS #LightYellow {
object "git" as git {
repo object DB
}
object "git-remote" as remote {
origin
}
object "git-tag" as tag {
v1.0.0
}
object "git-submodule" as submod {
lib/shared
}
object "git-branch" as branch {
main
}
object "git-stash" as stash {
stash@0
}
object "git-commit" as commit {
a1b2c3d
}
object "git-tree" as tree {
e8f1... root
}
object "git-tree-entry" as entry1 {
README.md
}
object "git-tree" as subtree {
src/ subtree
}
object "git-tree-entry" as entry2 {
src/app.ts
}
object "git-tree-entry" as entry3 {
src/main.ts
}
}
package "FILESYSTEM (worktree)" as FS #LightGreen {
object "fs-directory" as fsroot {
worktree root
}
object "fs-directory" as srcdir {
src/
}
object "fs-file" as readme {
README.md
}
object "fs-file" as app {
app.ts
}
object "fs-file" as main {
main.ts
}
}
package "STANDALONE FS-DIRECTORY" as SFS #Wheat {
object "fs-directory" as sfs {
/opt/deploy/myapp
}
object "fs-directory" as ssrcdir {
src/
}
object "fs-file" as sreadme {
README.md
}
object "fs-file" as sapp {
app.ts
}
object "fs-file" as smain {
main.ts
}
object "fs-symlink" as symlink {
link.txt
}
}
package "STANDALONE FS-MOUNT" as SM #LightCoral {
object "fs-mount" as mount {
/mnt/data
}
object "fs-directory" as mountroot {
root: /
}
}
package "VIRTUAL LAYER (abstract identities)" as VL #Lavender {
object "file" as vfile {
app.ts @ sha256:9f8e...
}
object "directory" as vdir {
src/ @ merkle:3d4f...
}
object "commit" as vcommit {
a1b2c3d
}
object "branch" as vbranch {
main @ a1b2c3d
}
object "remote" as vremote {
github.com/org/repo
}
object "tree" as vtree {
e8f1...9d2a
}
}
gc --> git
gc --> fsroot
mount --> mountroot
git --> remote
git --> tag
git --> submod
git --> branch
git --> stash
branch --> commit
commit --> tree
tree --> entry1
tree --> subtree
subtree --> entry2
subtree --> entry3
fsroot --> srcdir
fsroot --> readme
srcdir --> app
srcdir --> main
sfs --> ssrcdir
sfs --> sreadme
sfs --> symlink
ssrcdir --> sapp
ssrcdir --> smain
app ..> vfile : equivalent
sapp ..> vfile : equivalent
entry2 ..> vfile : equivalent
srcdir ..> vdir : equivalent
ssrcdir ..> vdir : equivalent
subtree ..> vdir : equivalent
commit ..> vcommit : equivalent
branch ..> vbranch : equivalent
remote ..> vremote : equivalent
tree ..> vtree : equivalent
@enduml
Reading the diagram:
- Top left: A
git-checkoutresource decomposes into agitchild (left — the repository's full structure) and anfs-directorychild (center — the worktree root directory at/home/user/projects/myapp). The worktree root is a directory, not a mount point. - Git structure (left column): The
gitchild containsgit-remote(origin),git-tag(v1.0.0),git-submodule(lib/shared),git-stash(stash@{0}), andgit-branch(main). The branch containsgit-commitobjects, each commit contains a rootgit-tree, and trees containgit-tree-entry(blobs) and nestedgit-tree(subtrees). This properly models git's internal object structure. - Filesystem (center/right): The worktree's
fs-directorycontainssrc/(anfs-directory),README.md(anfs-file), and files includingfs-symlinkresources. A standalonefs-directory(/opt/deploy/myapp) with the same structure. A standalonefs-mount(/mnt/data) with a rootfs-directorychild. - Virtual layer (boxed area): Shows how virtual types link equivalent physical resources:
- A
filevirtual resource linksfs-fileandgit-tree-entryresources that have the same content, filename, and permissions — bridging the filesystem and git layers. - A
directoryvirtual resource linksfs-directoryandgit-treeresources with the same recursive content. - A
commitvirtual resource linksgit-commitresources across repos with the same commit hash. - A
branchvirtual resource linksgit-branchresources across repos with the same name and HEAD. - A
remotevirtual resource linksgit-remoteresources across repos with the same URL. - A
treevirtual resource linksgit-treeresources across repos/commits with the same tree hash.
- A
Key separations:
gitvsgit-checkout: Agitresource represents a specific repository instance (local or remote). It can exist without any local directory (created from a remote URL — the repo exists on the remote server). Agit-checkoutalways has both agitchild and anfs-directorychild — it represents a locally cloned repo with files on disk.git-tree-entryvsfs-file: Agit-tree-entryis a blob entry in git's tree (path + content hash + mode). Anfs-fileis a physical file on disk. When a repo is checked out, both exist and typically have matching content — thefilevirtual type links them when content, name, and permissions match.git-treevsfs-directory: Agit-treeis a tree object in git's object database (a directory listing at a commit). Anfs-directoryis a physical directory on disk. Thedirectoryvirtual type links them when their recursive contents match.git-commit→git-tree→git-tree-entry: This chain properly models git internals. Commits point to a root tree, trees contain entries (blobs) and subtrees. The old design skipped the tree level; the current design preserves it.fs-mountvsfs-directory: Anfs-mountis a mount point (e.g.,/mnt/data). The filesystem type (ext4, btrfs, etc.) is a property. The root directory is anfs-directorychild. Anfs-directoryis a directory — it can be a child offs-mount,git-checkout, or anotherfs-directory, or it can be registered standalone by the user.git-checkoutcomposes withfs-directory, notfs-mount: A git checkout's worktree is a directory on an existing filesystem, not a separate mount point. This is whygit-checkout's required child isfs-directory(the worktree root), notfs-mount.- Virtual types bridge physical equivalents:
filebridgesfs-file+git-tree-entry(cross-layer).directorybridgesfs-directory+git-tree(cross-layer).commit,branch,tag,remote,submodule, andtreelink the same git structural element across different repos.
Concrete Example: A Made-Up Project
Consider a web application called "Acme Dashboard" with three registered resources:
local/acme-app— a checked-out git repo at/home/alice/projects/acme-dashboard(type:git-checkout)local/acme-upstream— a git repo accessed via remote URL, not cloned (type:git)local/acme-deploy— a standalone directory at/opt/deploy/acme-dashboardcontaining a production build snapshot (type:fs-directory)
Registration:
# 1) Checked-out git repo (has local files on disk) agents resource add git-checkout local/acme-app \ --path /home/alice/projects/acme-dashboard --branch main# 2) Git repo via remote URL (no local checkout — metadata only) agents resource add git local/acme-upstream </span> --url git@github.com:acmecorp/dashboard.git
# 3) Standalone directory (not a git repo — just files on disk) agents resource add fs-directory local/acme-deploy </span> --path /opt/deploy/acme-dashboard
What gets auto-discovered for each:
local/acme-app (type: git-checkout) discovers two children — a git and an fs-directory (the worktree root):
@startwbs
* local/acme-app\n(git-checkout / physical)
** local/acme-app:repo\n(git / physical)
*** acme-app:repo:origin\n(git-remote)\ngit@github.com:acmecorp/dashboard.git
*** acme-app:repo:v1.0.0\n(git-tag)
*** acme-app:repo:lib/shared\n(git-submodule @ c4d5e6f)
*** acme-app:repo:stash@0\n(git-stash)
*** acme-app:repo:main\n(git-branch)
**** main:a7f3e21\n(git-commit)
***** main:a7f3e21:tree\n(git-tree / root)
****** a7f3e21:README.md\n(git-tree-entry)
****** a7f3e21:package.json\n(git-tree-entry)
****** a7f3e21:src/\n(git-tree / subtree)
******* a7f3e21:src/app.ts\n(git-tree-entry)
******* a7f3e21:src/api.ts\n(git-tree-entry)
******* a7f3e21:src/utils.ts\n(git-tree-entry)
*** acme-app:repo:develop\n(git-branch)
**** develop:b2c4d8e\n(git-commit)
***** develop:b2c4d8e:tree\n(git-tree)
****** b2c4d8e:src/\n(git-tree)
******* b2c4d8e:src/app.ts\n(git-tree-entry)
******* b2c4d8e:src/api.ts\n(git-tree-entry / modified)
** local/acme-app:worktree\n(fs-directory / physical)\n/home/alice/projects/acme-dashboard/
*** worktree:src/\n(fs-directory)
**** worktree:src/app.ts\n(fs-file)
**** worktree:src/api.ts\n(fs-file)
**** worktree:src/utils.ts\n(fs-file)
*** worktree:package.json\n(fs-file)
*** worktree:README.md\n(fs-file)
*** worktree:docs -> ../docs\n(fs-symlink)
@endwbs
The git-checkout cleanly separates two concerns: the git child contains version control structure (remotes, branches, tags, stashes, submodules, commits, trees, and tree entries — git's full object model), while the fs-directory child is the worktree root directory containing the actual files on disk. Note how git's internal structure is fully modeled: git-commit → git-tree (root tree object) → git-tree-entry (blobs) and nested git-tree (subtrees). The worktree root is a directory (fs-directory), not a mount point — git-checkout does not own an fs-mount resource because a git checkout's worktree is just a directory on an existing filesystem. When content matches (as it does for a clean checkout), virtual types link the fs-file and git-tree-entry resources.
local/acme-upstream (type: git, remote URL — NOT checked out) discovers:
@startwbs
* local/acme-upstream\n(git / physical)\ngit@github.com:acmecorp/dashboard.git
** acme-upstream:origin\n(git-remote)
** acme-upstream:v1.0.0\n(git-tag)
** acme-upstream:main\n(git-branch)
*** main:a7f3e21\n(git-commit)
**** main:a7f3e21:tree\n(git-tree)
***** a7f3e21:src/\n(git-tree)
****** a7f3e21:src/app.ts\n(git-tree-entry)
****** a7f3e21:src/api.ts\n(git-tree-entry)
** acme-upstream:develop\n(git-branch)
*** ...\n(remaining structure)
@endwbs
A standalone git resource has full access to branches, tags, commits, trees, and tree entries — everything in the git object database — but no fs-directory child and no fs-file resources. There are no files on the local disk (the repo exists on GitHub's servers). Tools that need local file access cannot bind to it.
This is the key difference from git-checkout: a git resource represents a specific repo instance. Plans can reason about history, diffs between branches, remote relationships — without a local clone. A git-checkout adds the local worktree directory on top.
local/acme-deploy (type: fs-directory, standalone) discovers:
@startwbs
* local/acme-deploy\n(fs-directory / physical)\n/opt/deploy/acme-dashboard/
** acme-deploy:src/\n(fs-directory)
*** acme-deploy:src/app.ts\n(fs-file)
*** acme-deploy:src/api.ts\n(fs-file)
*** acme-deploy:src/utils.ts\n(fs-file)
** acme-deploy:package.json\n(fs-file)
** acme-deploy:README.md\n(fs-file)
@endwbs
A standalone fs-directory — no git metadata, no branches, no commits, no tree entries. Just a directory containing files and subdirectories. Uses the same fs-directory and fs-file types as the git checkout's worktree root.
Virtual resource linking across all three:
After all three resources are registered, the system detects equivalent physical resources and creates virtual parents to link them:
@startuml
skinparam defaultFontSize 10
skinparam objectFontSize 10
skinparam packageStyle rectangle
left to right direction
package "PHYSICAL: local/acme-app" as P1 #LightBlue {
object "fs-directory" as acmeWtSrc {
worktree:src/
}
object "fs-file" as acmeWtApp {
worktree:src/app.ts
}
object "fs-file" as acmeWtUtils {
worktree:src/utils.ts
}
object "fs-file" as acmeWtApi {
worktree:src/api.ts
}
object "git-tree" as acmeGitSrc {
main:a7f3e21:src/
}
object "git-tree-entry" as acmeGitApp {
main:src/app.ts
}
object "git-tree-entry" as acmeGitUtils {
main:src/utils.ts
}
object "git-tree-entry" as acmeGitApi {
main:src/api.ts
}
object "git-commit" as acmeCommit {
main:a7f3e21
}
object "git-branch" as acmeBranch {
main
}
object "git-tag" as acmeTag {
v1.0.0
}
object "git-remote" as acmeRemote {
origin
}
object "git-tree" as acmeTree {
main:a7f3e21:tree
}
object "git-submodule" as acmeSubmod {
lib/shared
}
}
package "PHYSICAL: local/acme-deploy" as P2 #LightGreen {
object "fs-directory" as deploySrc {
src/
}
object "fs-file" as deployApp {
src/app.ts
}
object "fs-file" as deployUtils {
src/utils.ts
}
object "fs-file" as deployApi {
src/api.ts
}
}
package "PHYSICAL: local/acme-upstream" as P3 #LightYellow {
object "git-tree" as upstreamSrc {
main:a7f3e21:src/
}
object "git-tree-entry" as upstreamApp {
main:src/app.ts
}
object "git-tree-entry" as upstreamUtils {
main:src/utils.ts
}
object "git-tree-entry" as upstreamApi {
main:src/api.ts
}
object "git-commit" as upstreamCommit {
main:a7f3e21
}
object "git-branch" as upstreamBranch {
main
}
object "git-tag" as upstreamTag {
v1.0.0
}
object "git-remote" as upstreamRemote {
origin
}
object "git-tree" as upstreamTree {
main:a7f3e21:tree
}
}
package "VIRTUAL LAYER\n(auto-created by equivalence)" as VL #Lavender {
object "directory" as vDir {
src/ (merkle:3d4f...)
}
object "file" as vAppTs {
app.ts (sha256:9f8e...)
}
object "file" as vUtilsTs {
utils.ts (sha256:a2b1...)
}
object "file" as vApiTs {
api.ts (sha256:e1d3...)
}
object "commit" as vCommit {
a7f3e21
}
object "branch" as vBranch {
main @ a7f3e21
}
object "tag" as vTag {
v1.0.0
}
object "remote" as vRemote {
github.com:acmecorp/dashboard.git
}
object "tree" as vTree {
e8f1...9d2a
}
object "submodule" as vSubmod {
lib/shared
}
}
acmeWtSrc ..> vDir
deploySrc ..> vDir
acmeGitSrc ..> vDir
upstreamSrc ..> vDir
acmeWtApp ..> vAppTs
deployApp ..> vAppTs
acmeGitApp ..> vAppTs
upstreamApp ..> vAppTs
acmeWtUtils ..> vUtilsTs
deployUtils ..> vUtilsTs
acmeGitUtils ..> vUtilsTs
upstreamUtils ..> vUtilsTs
acmeWtApi ..> vApiTs
deployApi ..> vApiTs
acmeGitApi ..> vApiTs
upstreamApi ..> vApiTs
note "NOT linked: develop:b2c4d8e:src/api.ts\n(different content on develop branch)" as N1
acmeCommit ..> vCommit
upstreamCommit ..> vCommit
acmeBranch ..> vBranch
upstreamBranch ..> vBranch
acmeTag ..> vTag
upstreamTag ..> vTag
acmeRemote ..> vRemote
upstreamRemote ..> vRemote
acmeTree ..> vTree
upstreamTree ..> vTree
acmeSubmod ..> vSubmod
@enduml
How the directory virtual type works:
The directory virtual resource for src/ exists because the src/ directory has identical recursive content across multiple physical locations. Its children include both fs-directory resources (physical directories on disk) and git-tree resources (tree objects in git's object database). The system detects directory equivalence by computing a Merkle hash over the sorted child content hashes. If someone adds a file to the deploy directory but not the git checkout's worktree, the Merkle hashes diverge, and local/acme-deploy:src/ is unlinked from the directory virtual parent.
How the file virtual type works:
The file virtual type links physical resources that represent the same file — identical content bytes, filename, and permissions. It bridges across layers: an fs-file on disk and a git-tree-entry in git's tree are linked when they have matching content. This is the primary cross-layer bridge. When content diverges (e.g., an uncommitted edit), the virtual link is broken.
How commit works across repos:
When two git resources share history (e.g., a fork and its upstream), the same commit hash will appear in both repos' branches. The commit virtual type links these — the commit object a7f3e21 in the local repo and a7f3e21 in the upstream repo are the same commit, and the virtual parent captures this identity.
Divergence scenario:
If a developer edits src/api.ts in the deploy directory (/opt/deploy/acme-dashboard/src/api.ts), the system detects the content hash change and:
- Unlinks
local/acme-deploy:src/api.tsfromfile: api.ts (sha256:e1d3...8a9b)(content no longer matches). - Unlinks
local/acme-deploy:src/fromdirectory: src/ (merkle:3d4f...a2b1)(directory contents no longer identical — the Merkle hash has changed). - The git checkout's worktree files and git-tree-entry resources remain linked (their content hasn't changed).
- If the edit makes the deploy file match some other known content hash, a new virtual link may be created.
Additional resource types (databases, APIs, cloud infrastructure, etc.) can be added as custom resource types via agents resource type add.
Custom Resource Types
Custom resource types are defined in YAML configuration files and registered via agents resource type add. Once registered, a custom type automatically becomes available as a new subcommand under agents resource add.
# File: resource-types/database.yaml cleveragents: version: "3.0"resource_type: name: local/database description: "A SQL database (PostgreSQL, MySQL, SQLite, etc.)" physical_or_virtual: physical user_addable: true
# CLI arguments for
agents resource add local/databasecli_arguments: - name: connection-string type: string required: true description: "Database connection string (e.g., postgresql://host/dbname)" validation: pattern: "^(postgresql|mysql|sqlite)://" - name: schema type: string required: false description: "Default schema to use" - name: read-only type: boolean required: false default: false description: "Whether the database should be treated as read-only"# Sandbox and handler sandbox_strategy: transaction_rollback handler: DatabaseHandler checkpointable: true
# Allowed parent types (empty means can be top-level) allowed_parent_types: []
# Child types child_types: - type: local/db-schema auto_discover: true manual_link: false description: "Discovered database schemas" - type: local/db-table auto_discover: true manual_link: false description: "Discovered tables within schemas"
# Capabilities capabilities: readable: true writable: true sandboxable: true checkpointable: true
When this type is registered:
agents resource type add --config ./resource-types/database.yaml
# Now available: agents resource add local/database <NAME> --connection-string CONN [--schema SCHEMA] [--read-only]
The user_addable field determines whether the type appears as a subcommand. Types with user_addable: false are only auto-generated as children — for example, git-remote, git-branch, git-commit, and git-tree-entry are never created directly by users but are discovered when a git (or git-checkout) resource is registered.
The Resource DAG
Resources form a directed acyclic graph (DAG), not a simple tree. A resource can have multiple parents and multiple children, subject to type constraints. The diagram below shows how a git-checkout, a standalone git repo (remote), a standalone fs-directory, and virtual resources interconnect:
@startuml
skinparam defaultFontSize 10
skinparam objectFontSize 10
skinparam packageStyle rectangle
package "GIT-CHECKOUT: local/app" as GCO #LightBlue {
object "git-checkout" as gco {
local/app
}
object "git" as gcoGit {
local/app:repo
}
object "git-remote" as gcoRemote {
origin
}
object "git-tag" as gcoTag {
v1.0.0
}
object "git-branch" as gcoBranch {
main
}
object "git-commit" as gcoCommit {
a1b2c3d
}
object "git-tree" as gcoTree {
e8f1... root
}
object "git-tree-entry" as gcoEntry1 {
README.md
}
object "git-tree" as gcoSubtree {
src/
}
object "git-tree-entry" as gcoEntry2 {
src/app.ts
}
object "git-tree-entry" as gcoEntry3 {
src/main.ts
}
object "fs-directory" as gcoFs {
local/app:worktree
}
object "fs-directory" as gcoSrcDir {
src/
}
object "fs-file" as gcoReadme {
README.md
}
object "fs-file" as gcoApp {
app.ts
}
object "fs-file" as gcoMain {
main.ts
}
}
package "GIT (remote): local/upstream" as GR #LightYellow {
object "git" as gr {
local/upstream
}
object "git-remote" as grRemote {
origin
}
object "git-tag" as grTag {
v1.0.0
}
object "git-branch" as grBranch {
main
}
object "git-commit" as grCommit {
a1b2c3d
}
object "git-tree" as grTree {
e8f1... root
}
object "git-tree-entry" as grEntry1 {
README.md
}
object "git-tree" as grSubtree {
src/
}
object "git-tree-entry" as grEntry2 {
src/app.ts
}
object "git-tree-entry" as grEntry3 {
src/main.ts
}
}
package "STANDALONE FS-DIRECTORY: local/deploy" as SFD #LightGreen {
object "fs-directory" as sfd {
/opt/deploy/myapp
}
object "fs-directory" as sfdSrcDir {
src/
}
object "fs-file" as sfdReadme {
README.md
}
object "fs-file" as sfdApp {
app.ts
}
object "fs-file" as sfdMain {
main.ts
}
}
package "VIRTUAL LAYER" as VL #Lavender {
object "file" as vFile {
app.ts@v1
}
object "directory" as vDir {
src/@v1
}
object "commit" as vCommit {
a1b2c3d
}
object "branch" as vBranch {
main@a1b...
}
object "tag" as vTag {
v1.0.0
}
object "remote" as vRemote {
github.com/org/repo
}
object "tree" as vTree {
e8f1...9d2a
}
}
' Physical hierarchy
gco --> gcoGit
gco --> gcoFs
gcoGit --> gcoRemote
gcoGit --> gcoTag
gcoGit --> gcoBranch
gcoBranch --> gcoCommit
gcoCommit --> gcoTree
gcoTree --> gcoEntry1
gcoTree --> gcoSubtree
gcoSubtree --> gcoEntry2
gcoSubtree --> gcoEntry3
gcoFs --> gcoSrcDir
gcoFs --> gcoReadme
gcoSrcDir --> gcoApp
gcoSrcDir --> gcoMain
gr --> grRemote
gr --> grTag
gr --> grBranch
grBranch --> grCommit
grCommit --> grTree
grTree --> grEntry1
grTree --> grSubtree
grSubtree --> grEntry2
grSubtree --> grEntry3
sfd --> sfdSrcDir
sfd --> sfdReadme
sfdSrcDir --> sfdApp
sfdSrcDir --> sfdMain
' Virtual equivalence links
gcoApp ..> vFile
sfdApp ..> vFile
gcoEntry2 ..> vFile
grEntry2 ..> vFile
gcoSrcDir ..> vDir
sfdSrcDir ..> vDir
gcoSubtree ..> vDir
grSubtree ..> vDir
gcoCommit ..> vCommit
grCommit ..> vCommit
gcoBranch ..> vBranch
grBranch ..> vBranch
gcoTag ..> vTag
grTag ..> vTag
gcoRemote ..> vRemote
grRemote ..> vRemote
gcoTree ..> vTree
grTree ..> vTree
@enduml
Key properties of the DAG:
- Multiple parents: A
git-tree-entryhas agit-treeparent (git structure) and potentially afilevirtual parent (content identity). Anfs-filehas anfs-directoryparent (filesystem hierarchy) and potentially afilevirtual parent. Anfs-directorycan be a child ofgit-checkout(as its worktree root), a child offs-mount(as the mount root), a child of anotherfs-directory(as a subdirectory), a child of adirectoryvirtual (identity), or a standalone user-registered resource. - Multiple children: A
git-checkouthas agitchild and anfs-directorychild. Agitresource hasgit-remote,git-branch,git-tag,git-stash,git-submodule, andgit-commitchildren. Agit-commithas agit-treechild. Agit-treehasgit-tree-entryand nestedgit-treechildren. - Cycles are forbidden: The graph is always a DAG. The system validates this when links are created.
- Type constraints: Not any resource can be a child of any other — the parent's resource type defines which child types are allowed (see the Allowed Parents / Allowed Children columns in the built-in types tables).
- Cross-layer bridge: Virtual resources link equivalent physical resources from different layers (git structure, filesystem, different repos). The
filevirtual type bridgesfs-file+git-tree-entry. Thedirectoryvirtual type bridgesfs-directory+git-tree. Thecommit,branch,tag,remote,submodule, andtreevirtual types link the same git structural element across different repositories.
Physical vs Virtual Resources
Resources are either physical or virtual, a distinction determined by their resource type.
Physical resources are specific, concrete manifestations. Each physical resource is a particular instance that exists somewhere — a file at a particular path on a particular machine, a git repository at a particular URL on a particular server, a commit in a particular repo. Physical resources can be directly read and written by tools. Most resources are physical.
Virtual resources represent an abstract identity that links equivalent physical resources. They answer the question: "where else does this same thing exist?" A file virtual resource links all the physical fs-file and git-tree-entry resources that have the same content, filename, and permissions. A commit virtual resource links git-commit resources across repos that share the same commit hash. Virtual resources have no location of their own and cannot be directly read or written.
Rules for the physical/virtual boundary:
- The children of a virtual resource can be physical resources, other virtual resources, or a combination of both.
- A physical resource's parents can be either physical or virtual resources.
- Not all physical resources need a virtual parent. Virtual resource linking is optional and only applies when equivalence tracking is meaningful.
Equivalence linking:
Two physical resources share a virtual parent when they are equivalent by the virtual type's criteria. Each virtual type has its own equivalence semantics:
-
file: Physical files (fs-fileorgit-tree-entry) share afileparent when they have the same content bytes (SHA-256), the same filename, and the same permissions. This is the primary cross-layer bridge — it links filesystem files with git blob entries. Example:local/app:worktree:src/main.py(anfs-file) andlocal/app:repo:main:a1b...:src/main.py(agit-tree-entry) both have the same content in a clean checkout, so they share afilevirtual parent. -
directory: Physical directories (fs-directory) and git trees (git-tree) share adirectoryparent when their full recursive contents are equivalent. This bridges the filesystem and git layers — thesrc/directory on disk and thesrc/subtree in git's tree object can be linked when their contents match. -
symlink: Physical symlinks (fs-symlink) and git tree entries with symlink mode (git-tree-entrymode 120000) share asymlinkparent when they have the same name and target. -
commit: Physical commits (git-commit) in different repos share acommitparent when they have the same commit hash. This is common when repos share history (e.g., a fork and its upstream). -
branch: Physical branches (git-branch) in different repos share abranchparent when they have the same name and the same HEAD commit hash. A push or local commit on one repo causes divergence. -
tag: Physical tags (git-tag) in different repos share atagparent when they have the same tag name and target object. -
remote: Physical remotes (git-remote) in different repos share aremoteparent when they point to the same URL. Answers: "these repos share the same upstream." -
submodule: Physical submodules (git-submodule) in different repos share asubmoduleparent when they have the same submodule URL and path. -
tree: Physical tree objects (git-tree) in different commits or repos share atreeparent when they have the same tree hash — identical directory structure and contents.
Divergence detection:
When a physical resource's content changes (e.g., a file is edited, a directory gains a new file, a branch advances), the system detects that it may no longer match its virtual parent's identity. At that point:
- If the edit causes the physical resource to diverge from all other physical siblings under the same virtual parent, the physical resource is unlinked from that virtual parent.
- If the edit makes it match a different virtual resource's identity, it may be re-linked.
- Content equivalence is tracked via content hashing: SHA-256 for files, Merkle hashes for directories, tree hashes for git trees, commit hashes for commits, HEAD+name for branches, name+target for tags, URL for remotes.
- Cascading divergence: When a
filelink breaks, the parentdirectoryvirtual resource is also re-evaluated (since its identity depends on all child identities).
This enables powerful queries like:
- "What physical locations exist for this file content?" → Find all physical resources sharing the
filevirtual parent. - "What tools can edit this file?" → Find all tools with resource bindings compatible with the physical resource types.
- "Has this file been modified in any location?" → Check if all physical siblings still share the same
filevirtual parent. - "Are these two repos in sync?" → Check if their branches share a
branchvirtual parent. - "Which directories are identical across deployments?" → Find
directoryvirtual resources with multiple physical children. - "What submodules are shared across projects?" → Find
submodulevirtual resources with multiple physical children.
Resource Registration (CLI)
Resources are created via agents resource add <type> with type-specific arguments. Auto-discovered children are created automatically:
# Register a checked-out git repository (most common) agents resource add git-checkout local/api-repo --path /home/user/projects/api-service --branch main # Auto-discovers: git child (repo metadata, remotes, branches, commits, tree entries) # + fs-directory child (worktree root directory, subdirectories, files)# Register a git repo via remote URL (no local checkout — exists on remote server) agents resource add git local/upstream --url git@github.com:org/upstream.git # Auto-discovers: remotes, branches, tags, commits, trees, tree entries, stashes, submodules # (no fs-directory — not checked out locally)
# Register a standalone directory (not a git repo — just files on disk) agents resource add fs-directory local/docs --path /opt/docs/api-reference # Auto-discovers: subdirectories, files, symlinks, hardlinks
# Register a standalone filesystem mount (entire volume) agents resource add fs-mount local/data-volume --mount-path /mnt/data # Auto-discovers: root fs-directory, subdirectories, files, symlinks, hardlinks
# Link resources to projects (resources must be registered first) agents project link-resource local/api-service local/api-repo agents project link-resource local/api-service local/docs --read-only
Auto-Discovery
When a resource is registered, its resource type's handler auto-discovers child resources. This process:
- Scans the resource to identify children (e.g., a
git-checkouthandler creates agitchild and anfs-directorychild for the worktree root; agithandler lists remotes, branches, tags, commits, stashes, and submodules; agit-commithandler creates a rootgit-treechild; agit-treehandler lists tree entries and subtrees; anfs-mounthandler creates a rootfs-directoryand discovers its contents; anfs-directoryhandler discovers subdirectories, files, symlinks, and hardlinks). - Creates child resource records in the Resource Registry identified by auto-generated ULIDs. Auto-discovered children do not receive names — they are identified by ULID only, even if their resource type is user-addable.
- Reuses existing resources: If a discovered child matches an already-registered resource (same type + same value/location), the existing resource is linked as a child rather than creating a duplicate. For example, if
local/docs(anfs-directoryresource at/home/user/projects/api-service) is already registered and thegit-checkouthandler discovers its worktree root at the same path, it links to the existinglocal/docsresource instead of creating a new one. - Links virtual resources: When auto-discovery detects that a physical resource's content matches an existing virtual resource (via content hashing), it links the physical resource as a child of the virtual resource.
Auto-discovered children are marked as auto: true in the DAG, distinguishing them from manually linked children. Auto-discovered links cannot be manually unlinked (they are managed by the handler), while manually created links can be freely managed.
Auto-discovery runs:
- At registration time (
agents resource add) - On refresh (when the system detects changes, e.g., new commits, new files)
- On demand (when a tool accesses the resource and the handler detects staleness)
Resource Capabilities
Each resource declares its capabilities, derived from its resource type:
| Capability | Description |
|---|---|
readable |
Whether the resource can be read |
writable |
Whether the resource can be modified |
sandboxable |
Whether the resource supports sandbox isolation |
checkpointable |
Whether the resource supports checkpoint/rollback |
These capabilities are used by the tool execution flow to validate that a tool's resource binding is compatible — a tool declaring access: read_write on a resource slot cannot be bound to a read-only resource.
Resource Registry
CleverAgents maintains a Resource Registry — a persistent catalog of all registered resources and their DAG relationships:
@startuml
skinparam classAttributeIconSize 0
skinparam classFontSize 13
skinparam defaultFontSize 12
class ResourceRegistry {
- resourceIndex : Map<String, ResourceRecord>
- typeIndex : Map<String, ResourceTypeRecord>
--
+ add(type, name, properties) : ResourceRecord
+ update(name, properties) : ResourceRecord
+ remove(name) : void
+ lookup(name) : ResourceRecord
+ list(filters) : ResourceRecord[]
+ tree(name, depth) : DAG_subtree
+ link_child(parent, child) : void
+ unlink_child(parent, child) : void
+ refresh(name) : void
+ find_by_content(hash) : ResourceRecord[]
+ find_virtual_parent(resource) : ResourceRecord
}
class ResourceRecord {
+ name : String
+ type : String
+ physical_or_virtual : PhysVirt
+ properties : Map<String, Object>
+ capabilities : Capabilities
+ parents : List<ResourceRecord>
+ children : List<ResourceRecord>
+ content_hash : String
+ linked_projects : List<String>
+ created_at : DateTime
+ updated_at : DateTime
}
class ResourceTypeRecord {
+ name : String
+ description : String
+ source : String
+ physical_or_virtual : PhysVirt
+ user_addable : Boolean
+ cli_arguments : List<CLIArgument>
+ allowed_parent_types : List<String>
+ child_types : Map<String, ChildTypeConfig>
+ sandbox_strategy : String
+ handler : String
+ capabilities : Capabilities
+ config_path : String
}
enum PhysVirt {
physical
virtual
}
ResourceRegistry "1" *-- "0..*" ResourceRecord : indexes resources >
ResourceRegistry "1" *-- "0..*" ResourceTypeRecord : indexes types >
ResourceRecord --> PhysVirt
ResourceTypeRecord --> PhysVirt
note right of ResourceRegistry
**Populated by:**
- agents resource add CLI
- Auto-discovery during registration
- Content-identity linking
**Consumed by:**
- Tool binding resolution
- Sandbox creation
- Change tracking
- Project linking
- Plan validation
end note
@enduml
The Resource Registry persists in the database (local SQLite or server). It works alongside the Tool Registry and Skill Registry.
Resource Sandbox Strategy
Each resource defines its own sandbox strategy, determined by its resource type. This is critical because:
- The same resource may be accessed through different tools and skills
- Different resource types require different sandboxing approaches
- Some resources cannot be sandboxed at all
| Resource Type | Sandbox Strategy | Rollback Mechanism |
|---|---|---|
git-checkout |
git_worktree |
Git reset/checkout |
git |
none |
N/A (represents a repo instance — not directly sandboxable) |
fs-mount |
copy_on_write or overlay |
Restore from snapshot |
fs-directory |
copy_on_write |
Restore from snapshot |
| Custom database types | transaction_rollback |
Transaction rollback |
| Custom API types | none (often not sandboxable) |
N/A |
The sandbox strategy is inherited by child resources from their parent unless the child type defines its own. For example, git-branch, git-commit, git-tree, and git-tree-entry all inherit from their git ancestor. fs-file, fs-symlink, and fs-hardlink inherit from their fs-directory parent.
Lazy Sandboxing
Resources are sandboxed lazily when accessed, not upfront. This is different from indexing — resources are indexed immediately when registered, but sandboxes are only created when execution needs to modify a resource:
- A project may link many resources (e.g., git repo + databases + cloud accounts)
- A plan may only need to modify one resource
- Only the accessed resources are sandboxed
- Each plan/child plan has its own sandbox containing only edited resources
This is efficient for large projects where most resources remain untouched.
Resource Access Tracking
The system tracks which tools access which resources through the tool-resource binding system. This tracking happens at multiple levels:
- Declaration-time: Tool YAML declares resource slots with type and access mode (see Resource Bindings in the Tools section).
- Activation-time: When a tool is activated for a plan, its resource slots are bound to specific resources. The system records these bindings.
- Execution-time: Every tool invocation logs which bound resources were actually accessed, what operations were performed (read/write/delete), and what paths or objects were touched.
This enables:
- Accurate sandbox scoping: Only sandbox resources that will actually be modified.
- Rollback feasibility analysis: Know exactly which resources were modified and whether they support rollback.
- Security auditing: Complete record of which tools accessed which resources and how.
- Cross-resource impact analysis: Determine if changes to a resource affect tools bound to sibling or child resources.
- Virtual resource consistency: Detect when a physical resource diverges from its virtual parent.
Unified Resource Abstraction Layer
CleverAgents provides a unified abstraction that allows tools to work with any resource type through a consistent interface. This enables:
- Resource-agnostic tools: A tool like
read_content(path)works whether the path refers to a file, database record, or API endpoint. - Consistent sandbox semantics: All resources support the same sandbox lifecycle (create, read, write, checkpoint, rollback).
- Pluggable resource handlers: New resource types can be added by registering custom resource types without modifying existing tools.
- Unified change tracking: All resource modifications flow into the same ChangeSet model.
Resource Handler Interface
Every resource type provides a handler that implements this interface:
class ResourceHandler(Protocol): """Handler for a specific resource type."""<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">read</span>(self, path: <span style="color: cyan;">str</span>, sandbox: Sandbox) -> Content: <span style="color: #66cc66;">"""Read content from the sandboxed resource."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">write</span>(self, path: <span style="color: cyan;">str</span>, content: Content, sandbox: Sandbox) -> Change: <span style="color: #66cc66;">"""Write content and return the Change record."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">delete</span>(self, path: <span style="color: cyan;">str</span>, sandbox: Sandbox) -> Change: <span style="color: #66cc66;">"""Delete resource and return the Change record."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;"><span style="color: cyan;">list</span></span>(self, pattern: <span style="color: cyan;">str</span>, sandbox: Sandbox) -> <span style="color: cyan;">list</span>[<span style="color: cyan;">str</span>]: <span style="color: #66cc66;">"""List paths matching pattern."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">diff</span>(self, path: <span style="color: cyan;">str</span>, sandbox: Sandbox) -> <span style="color: cyan;">str</span>: <span style="color: #66cc66;">"""Generate diff between sandbox and original state."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">supports_operation</span>(self, operation: OperationType) -> <span style="color: cyan;">bool</span>: <span style="color: #66cc66;">"""Check if this resource supports the given operation."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">discover_children</span>(self, resource: ResourceRecord) -> <span style="color: cyan;">list</span>[ResourceRecord]: <span style="color: #66cc66;">"""Auto-discover child resources (called at registration and refresh)."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">content_hash</span>(self, path: <span style="color: cyan;">str</span>, sandbox: Sandbox) -> <span style="color: cyan;">str</span>: <span style="color: #66cc66;">"""Compute content hash for identity tracking."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">create_sandbox</span>(self, resource: ResourceRecord) -> Sandbox: <span style="color: #66cc66;">"""Create a sandbox for this resource using its type's strategy."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">create_checkpoint</span>(self, sandbox: Sandbox) -> Checkpoint: <span style="color: #66cc66;">"""Create a checkpoint within the sandbox."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">rollback_to</span>(self, sandbox: Sandbox, checkpoint: Checkpoint) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Roll back sandbox state to a checkpoint."""</span> ...
Built-in Resource Handlers
| Resource Type | Handler | Read | Write | Delete | Sandbox Strategy |
|---|---|---|---|---|---|
git-checkout |
GitCheckoutHandler |
✓ | ✓ | ✓ | git_worktree |
git |
GitHandler |
✓ | ✗ | ✗ | none |
git-commit, git-tree, git-tree-entry |
GitObjectHandler |
✓ | ✗ | ✗ | (inherits) |
git-branch, git-tag, git-stash |
GitRefHandler |
✓ | ✓ | ✓ | (inherits) |
git-remote, git-submodule |
GitConfigHandler |
✓ | ✗ | ✗ | (inherits) |
fs-mount |
FilesystemHandler |
✓ | ✓ | ✓ | copy_on_write |
fs-directory |
FilesystemHandler |
✓ | ✓ | ✓ | copy_on_write |
fs-file, fs-symlink, fs-hardlink |
FilesystemHandler |
✓ | ✓ | ✓ | (inherits) |
Additional handlers are provided by custom resource types when they are registered.
Resource Path Resolution
Paths in tool invocations are resolved through a resource routing system that uses tool-resource bindings to determine the target resource:
sequenceDiagram
participant Tool as Tool Invocation
participant Router as Resource Router
participant Handler as GitCheckoutHandler
Tool->>Router: edit_file(path='src/main.py', ...)
Router->>Router: Check tool's resource bindings
Note right of Router: slot 'repo' bound to<br/>local/api-repo (git-checkout)
Router->>Router: Resolve path within bound resource
Note right of Router: local/api-repo:worktree:src/main.py
Router->>Handler: Route to resource handler
Handler->>Handler: Resolve path to sandbox worktree
Handler->>Handler: Operate on sandboxed state
Handler-->>Tool: Return Change record
When a tool has multiple resource slots bound, the path scheme or slot name disambiguates:
# Explicit slot reference path://repo/src/main.py → routes to the "repo" slot's bound resource path://docs/api/readme.md → routes to the "docs" slot's bound resourceDefault: unqualified paths route to the tool's primary resource slot
src/main.py → routes to the first (or only) resource slot
Context
!!! adr "Architecture Decision" The Advanced Context Management System, UKO ontology, CRP protocol, and context strategies are defined in ADR-014: Context Management (ACMS).
The Advanced Context Management System (ACMS) is CleverAgents' fully pluggable, strategy-driven framework for assembling context for actors. It replaces and subsumes the basic context tier system with a comprehensive architecture built around three core innovations:
-
A Universal Knowledge Ontology (UKO) -- an inheritance-based RDF ontology hierarchy that represents any resource (source code, documents, databases, infrastructure) at multiple levels of abstraction, with full provenance back to the originating resource and temporal versioning across revisions.
-
A Context Request Protocol (CRP) -- a structured vocabulary through which actors (via skills) declare what information they need, at what level of detail, and with what scope, while they are reasoning. This replaces static "context view" approaches with a dynamic, demand-driven model.
-
A Context Assembly Pipeline -- a pluggable, 10-component pipeline that replaces the former monolithic Strategy Coordinator and Fusion Engine. Multiple independent context strategies can be registered, each working with different data backends and different abstraction levels of the UKO. The pipeline orchestrates strategy selection, budget allocation, parallel execution, fragment deduplication, depth resolution, scoring, budget packing, ordering, preamble generation, and skeleton compression -- each step backed by a replaceable Protocol implementation configurable at global, project, or plan scope.
The system is designed around the hierarchical nature of plans: parent plans see wide, shallow context (breadth); child plans see narrow, deep context (depth). Context inheritance flows parent-to-child with progressive focusing, and every context assembly respects a dynamically-changing token budget. This architecture is domain-agnostic: the same breadth/depth/DetailDepth mechanics work identically whether the underlying resources are source code repositories, technical documents, database schemas, or infrastructure configurations.
Critical Design Decision: All indexing happens immediately when resources are added to projects or when code changes. There is no "on-demand" indexing during agent execution. This ensures that agents always have instant access to search capabilities without any indexing delays. The computational cost is paid once upfront, not repeatedly during agent operations.
The full architectural design of the ACMS -- including the UKO ontology hierarchy, backend abstraction layer, Context Assembly Pipeline (the 10-component pluggable pipeline that replaced the former Strategy Coordinator and Fusion Engine), index synchronization, performance characteristics, and plugin architecture -- is specified in the Architecture > ACMS Architecture section.
Core Data Types
The ACMS defines several data types that are used throughout the system by actors, skills, strategies, and the plan lifecycle.
DetailDepth and DetailLevelMap
At the most general level (UKO Layer 0), detail depth is simply a non-negative integer — 0 being the most minimal representation and each increment revealing more. There is no fixed upper bound; the maximum meaningful depth depends on the domain and the complexity of the information unit. This is the DetailDepth type.
Each UKO domain extension then registers a DetailLevelMap — a table of named labels mapped to specific integer depths. This allows domain users to work with meaningful names (like MODULE_LISTING or TABLE_OF_CONTENTS) rather than raw integers, while the underlying system always operates on the universal integer scale. Maps are inherited: a language-specific map includes all levels from its parent paradigm map, which includes all levels from the general code map.
DetailDepth = int # Non-negative integer, 0 = most minimal, no upper bound@dataclass class DetailLevelMap: """Maps named detail levels to integer depths for a UKO domain. Inherited: a child map includes all entries from its parent map, and may insert additional levels at any integer position.""" domain: str # UKO namespace (e.g., "uko-code:", "uko-py:") parent: DetailLevelMap | None # Parent map to inherit from levels: dict[str, int] # Named level -> integer depth max_depth: int # Maximum meaningful depth for this domain
<span style="color: cyan; font-weight: 600;">def</span> <span style="color: #66cc66;">resolve</span>(<span style="color: #5599ff;">self</span>, depth: <span style="color: #66cc66;">int | str</span>) -> <span style="color: #66cc66;">int</span>: <span style="color: #888;">"""Resolve a named level or integer to an integer depth."""</span> <span style="color: cyan; font-weight: 600;">if</span> <span style="color: #66cc66;">isinstance</span>(depth, <span style="color: #66cc66;">int</span>): <span style="color: cyan; font-weight: 600;">return</span> <span style="color: #66cc66;">min</span>(depth, <span style="color: #5599ff;">self</span>.max_depth) <span style="color: #888;"># Look up named level in this map, then parent maps</span> <span style="color: cyan; font-weight: 600;">if</span> depth <span style="color: cyan; font-weight: 600;">in</span> <span style="color: #5599ff;">self</span>.levels: <span style="color: cyan; font-weight: 600;">return</span> <span style="color: #5599ff;">self</span>.levels[depth] <span style="color: cyan; font-weight: 600;">if</span> <span style="color: #5599ff;">self</span>.parent: <span style="color: cyan; font-weight: 600;">return</span> <span style="color: #5599ff;">self</span>.parent.resolve(depth) <span style="color: cyan; font-weight: 600;">raise</span> <span style="color: #ff6666;">ValueError</span>(<span style="color: #66cc66;">f"Unknown detail level: {depth}"</span>)
Universal (Layer 0) Semantics:
At the universal level, each integer depth has a domain-agnostic meaning. The principle is that depth 0 answers "what exists?", and each subsequent depth answers progressively more detailed questions:
| Depth | Universal Question Answered | What Gets Included |
|---|---|---|
| 0 | What exists? | Just the name/identifier of the information unit. |
| 1 | How is it organized? | Names of immediate children (structural skeleton). |
| 2 | What are the key relationships? | Children + dependency/reference edges to other units. |
| 3 | What is each thing's purpose? | + Short descriptions/summaries for each child. |
| 4 | What is the structural shape? | + Type information, size/count metadata, categories. |
| 5+ | How does it work? What does it say? | Progressively more content, up to complete verbatim content at max depth. |
| max | Everything. | Complete content — nothing omitted. |
The exact number of meaningful depths varies by domain (source code may have 12+ levels, a flat config file may only have 4). The system never forces content into a fixed number of buckets.
Source Code DetailLevelMap (uko-code:, extended by uko-oo:, uko-py:, etc.):
The general software domain defines a base set of named levels. Paradigm-specific and language-specific extensions insert additional levels where their semantics provide meaningful distinctions.
| Depth | uko-code: Name |
Content Shown |
|---|---|---|
| 0 | MODULE_LISTING |
Module/package names only, no internal structure. |
| 1 | MODULE_GRAPH |
Module names + inter-module dependency edges (imports graph). |
| 2 | MEMBER_LISTING |
+ Names of top-level members within each module (classes, functions, constants) — no signatures. |
| 3 | MEMBER_SUMMARY |
+ One-line docstring or LLM-generated summary for each member. |
| 4 | SIGNATURES |
+ Full type signatures (parameter names, types, return types) for all callables and type definitions. |
| 5 | SIGNATURES_WITH_DOCS |
+ Complete docstrings/comments attached to each member. |
| 6 | STRUCTURAL_OUTLINE |
+ Control flow structure within callable bodies (branches, loops, try/except) shown as outline, no expressions. |
| 7 | KEY_LOGIC |
+ Key expressions: return statements, assertions, assignments to important variables. |
| 8 | NEAR_COMPLETE |
+ All statements except comments, logging, and boilerplate (imports, __all__, etc.). |
| 9 | FULL_SOURCE |
Complete source code — nothing omitted. |
Object-Oriented effective map (uko-oo:) — inherits from uko-code:, inserts CLASS_HIERARCHY and VISIBILITY_ANNOTATED:
When the uko-oo: extension is active, the effective DetailLevelMap is re-numbered with consecutive integers. The two inserted levels shift all subsequent depths upward:
| Depth | Name | Origin | Content Shown |
|---|---|---|---|
| 0 | MODULE_LISTING |
uko-code: |
Module/package names only, no internal structure. |
| 1 | MODULE_GRAPH |
uko-code: |
Module names + inter-module dependency edges (imports graph). |
| 2 | MEMBER_LISTING |
uko-code: |
+ Names of top-level members within each module (classes, functions, constants) — no signatures. |
| 3 | CLASS_HIERARCHY |
uko-oo: |
Inheritance chains and interface implementation relationships between classes. |
| 4 | MEMBER_SUMMARY |
uko-code: |
+ One-line docstring or LLM-generated summary for each member. |
| 5 | SIGNATURES |
uko-code: |
+ Full type signatures (parameter names, types, return types) for all callables and type definitions. |
| 6 | SIGNATURES_WITH_DOCS |
uko-code: |
+ Complete docstrings/comments attached to each member. |
| 7 | VISIBILITY_ANNOTATED |
uko-oo: |
Public/protected/private modifiers on all members + abstract/final markers. |
| 8 | STRUCTURAL_OUTLINE |
uko-code: |
+ Control flow structure within callable bodies (branches, loops, try/except) shown as outline, no expressions. |
| 9 | KEY_LOGIC |
uko-code: |
+ Key expressions: return statements, assertions, assignments to important variables. |
| 10 | NEAR_COMPLETE |
uko-code: |
+ All statements except comments, logging, and boilerplate (imports, __all__, etc.). |
| 11 | FULL_SOURCE |
uko-code: |
Complete source code — nothing omitted. |
Python-specific effective map (uko-py:) — inherits from uko-oo:, inserts DECORATED_SIGNATURES, TYPE_STUBS, and appends WITH_TESTS:
When the uko-py: extension is active, the effective map builds on uko-oo: and re-numbers again:
| Depth | Name | Origin | Content Shown |
|---|---|---|---|
| 0 | MODULE_LISTING |
uko-code: |
Module/package names only, no internal structure. |
| 1 | MODULE_GRAPH |
uko-code: |
Module names + inter-module dependency edges (imports graph). |
| 2 | MEMBER_LISTING |
uko-code: |
+ Names of top-level members within each module (classes, functions, constants) — no signatures. |
| 3 | CLASS_HIERARCHY |
uko-oo: |
Inheritance chains and interface implementation relationships between classes. |
| 4 | MEMBER_SUMMARY |
uko-code: |
+ One-line docstring or LLM-generated summary for each member. |
| 5 | SIGNATURES |
uko-code: |
+ Full type signatures (parameter names, types, return types) for all callables and type definitions. |
| 6 | SIGNATURES_WITH_DOCS |
uko-code: |
+ Complete docstrings/comments attached to each member. |
| 7 | DECORATED_SIGNATURES |
uko-py: |
@property, @staticmethod, @classmethod, custom decorator chains shown on each member. |
| 8 | VISIBILITY_ANNOTATED |
uko-oo: |
Public/protected/private modifiers on all members + abstract/final markers. |
| 9 | STRUCTURAL_OUTLINE |
uko-code: |
+ Control flow structure within callable bodies (branches, loops, try/except) shown as outline, no expressions. |
| 10 | KEY_LOGIC |
uko-code: |
+ Key expressions: return statements, assertions, assignments to important variables. |
| 11 | TYPE_STUBS |
uko-py: |
Type annotations from .pyi stub files merged with source, typing module usage. |
| 12 | NEAR_COMPLETE |
uko-code: |
+ All statements except comments, logging, and boilerplate (imports, __all__, etc.). |
| 13 | FULL_SOURCE |
uko-code: |
Complete source code — nothing omitted. |
| 14 | WITH_TESTS |
uko-py: |
Full source + associated test cases for each callable (from uko-code:testsCallable edges). |
Document DetailLevelMap (uko-doc:):
| Depth | uko-doc: Name |
Content Shown |
|---|---|---|
| 0 | TITLE_ONLY |
Document title / section heading only. |
| 1 | TABLE_OF_CONTENTS_L1 |
Title + top-level (depth-1) section/chapter headings. |
| 2 | TABLE_OF_CONTENTS_L2 |
+ Second-level subsection headings. |
| 3 | FULL_TOC |
Complete table of contents to all heading depths. |
| 4 | TOC_WITH_SUMMARIES |
Full TOC + one-sentence abstract per section (LLM-generated or extracted from first paragraph). |
| 5 | SECTION_OVERVIEWS |
+ Topic keywords per section + cross-reference targets (which other sections this section discusses). |
| 6 | TOPIC_SENTENCES |
+ First sentence of every paragraph within each section. |
| 7 | PARAGRAPH_SUMMARIES |
+ LLM-generated summary for each paragraph (2-3 sentences compressed to 1). |
| 8 | STRUCTURAL_DETAIL |
+ All figure/table captions + code block headers + list item first lines + blockquote sources. |
| 9 | NEAR_COMPLETE |
+ Full paragraph text, but with inline formatting stripped and code blocks truncated to first/last 3 lines. |
| 10 | FULL_CONTENT |
Complete document content — all text, formatting, code blocks, figures, footnotes. |
Database DetailLevelMap (uko-data:):
| Depth | uko-data: Name |
Content Shown |
|---|---|---|
| 0 | SCHEMA_LISTING |
Schema/database names only. |
| 1 | TABLE_LISTING |
+ Table and view names within each schema. |
| 2 | COLUMN_LISTING |
+ Column names within each table (no types). |
| 3 | TYPED_COLUMNS |
+ Column data types + nullability. |
| 4 | CONSTRAINTS |
+ Primary keys, unique constraints, check constraints, defaults. |
| 5 | RELATIONSHIPS |
+ Foreign key relationships (full referential graph between tables). |
| 6 | INDEXES_AND_STATS |
+ Index definitions + estimated row counts + basic statistics (cardinality, avg row size). |
| 7 | DDL |
Full CREATE TABLE / CREATE VIEW DDL (reconstructed from metadata). |
| 8 | DDL_WITH_TRIGGERS |
+ Trigger definitions + stored procedure signatures that reference each table. |
| 9 | FULL_PROCEDURES |
+ Complete stored procedure and function bodies. |
| 10 | WITH_SAMPLE_DATA |
+ Sample rows (configurable N, default 5) for each table. |
| 11 | FULL_CATALOG |
Complete catalog: DDL + all procedure bodies + triggers + sample data + value distribution histograms + query plan statistics. |
Infrastructure DetailLevelMap (uko-infra:):
| Depth | uko-infra: Name |
Content Shown |
|---|---|---|
| 0 | SERVICE_LISTING |
Service/component names only. |
| 1 | SERVICE_GRAPH |
+ Service dependency edges (which services connect to which). |
| 2 | ENDPOINT_LISTING |
+ Endpoint paths/ports for each service. |
| 3 | ENDPOINT_DETAIL |
+ HTTP methods, parameters, authentication requirements per endpoint. |
| 4 | CONFIG_KEYS |
+ Configuration key names and their sections. |
| 5 | CONFIG_VALUES |
+ Configuration values (with secrets masked). |
| 6 | RESOURCE_LIMITS |
+ Resource allocations (CPU, memory, replicas, storage). |
| 7 | FULL_CONFIG |
Complete configuration files for each service. |
| 8 | WITH_DEPLOYMENT |
+ Deployment descriptors (Dockerfiles, Kubernetes manifests, Compose files). |
How depth resolution works in practice: When a context request specifies depth="SIGNATURES" and the target node is a uko-py:Module, the system looks up SIGNATURES in the uko-py: DetailLevelMap first. If found, it uses that integer. If not, it walks up to uko-oo:, then uko-code:, then uko: until it finds a match. If the request specifies depth=4 (a raw integer), it is used directly — the system renders the target node at integer depth 4 regardless of what named level that corresponds to.
ContextRequest
A structured request for context, issued by an actor or skill via the CRP.
@dataclass class ContextRequest: """A structured request for context, issued by an actor or skill."""<span style="color: #888;"># === What to find ===</span> query: <span style="color: cyan;">str</span> | <span style="color: cyan;">None</span> = <span style="color: magenta; font-weight: 600;">None</span> <span style="color: #888;"># Natural language query</span> entities: <span style="color: cyan;">list</span>[<span style="color: cyan;">str</span>] = field(default_factory=<span style="color: cyan;">list</span>) <span style="color: #888;"># Named entities to focus on</span> uko_types: <span style="color: cyan;">list</span>[<span style="color: cyan;">str</span>] = field(default_factory=<span style="color: cyan;">list</span>) <span style="color: #888;"># UKO types to filter</span> <span style="color: #888;"># === Scope control ===</span> focus: <span style="color: cyan;">list</span>[<span style="color: cyan;">str</span>] = field(default_factory=<span style="color: cyan;">list</span>) <span style="color: #888;"># URIs or identifiers of specific items to focus on</span> breadth: <span style="color: cyan;">int</span> = <span style="color: yellow;">2</span> <span style="color: #888;"># How many hops outward in the dependency/reference graph.</span> depth: <span style="color: cyan;">int</span> | <span style="color: cyan;">str</span> = <span style="color: yellow;">3</span> <span style="color: #888;"># How much detail to include for each item found.</span> <span style="color: #888;"># May be a raw integer (0-N) or a named level string (e.g., "SIGNATURES")</span> <span style="color: #888;"># resolved via the active DetailLevelMap for the target node's UKO type.</span> <span style="color: #888;"># === Focus depth gradient ===</span> depth_gradient: <span style="color: cyan;">bool</span> = <span style="color: magenta; font-weight: 600;">True</span> <span style="color: #888;"># When True, items closer to the focus get more detail.</span> <span style="color: #888;"># === Temporal scope ===</span> temporal: <span style="color: cyan; font-weight: 600;">TemporalScope</span> = <span style="color: cyan; font-weight: 600;">TemporalScope</span>.CURRENT <span style="color: #888;"># === Budget ===</span> max_tokens: <span style="color: cyan;">int</span> | <span style="color: cyan;">None</span> = <span style="color: magenta; font-weight: 600;">None</span> <span style="color: #888;"># === Strategy hints ===</span> preferred_strategies: <span style="color: cyan;">list</span>[<span style="color: cyan;">str</span>] = field(default_factory=<span style="color: cyan;">list</span>) required_backends: <span style="color: cyan;">list</span>[<span style="color: cyan;">str</span>] = field(default_factory=<span style="color: cyan;">list</span>) <span style="color: #888;"># === Priority and purpose ===</span> priority: <span style="color: cyan;">float</span> = <span style="color: yellow;">0.5</span> <span style="color: #888;"># 0.0 = background, 1.0 = critical</span> purpose: <span style="color: cyan;">str</span> = <span style="color: #66cc66;">""</span> <span style="color: #888;"># Why is this context needed?</span>
ContextFragment
The atomic unit of context returned by strategies and consumed by the Context Assembly Pipeline's fusion phase (FragmentDeduplicator, DetailDepthResolver, FragmentScorer, BudgetPacker, FragmentOrderer).
@dataclass
class ContextFragment:
"""A single piece of context assembled by a strategy."""
uko_node: str # UKO URI of the source node
content: str # Rendered text content
detail_depth: int # Resolved integer depth of this fragment
token_count: int # Token count of content
relevance_score: float # 0.0-1.0 relevance to the request
provenance: FragmentProvenance # Trace back to resource + location
metadata: dict = field(default_factory=dict)
AssembledContext
The output of a context assembly cycle -- the final, budget-respecting payload delivered to an actor.
@dataclass
class AssembledContext:
"""The fused, budget-respecting context payload."""
fragments: list[ContextFragment] # Ordered context fragments
total_tokens: int # Total token count
budget_used: float # Fraction of budget consumed (0.0-1.0)
strategies_used: list[str] # Which strategies contributed
context_hash: str # Cryptographic hash for snapshot
preamble: str | None # Optional structure summary
provenance_map: dict # Fragment -> resource/location mapping
Context Request Protocol (CRP)
Instead of static context views, actors (through their skills) actively request context as they reason. The CRP defines a structured vocabulary for these requests.
This is fundamentally different from existing approaches where context is assembled once before the actor runs. With CRP:
- The actor starts with an initial context (assembled from the plan's context view, parent context inheritance, and strategy defaults).
- As the actor reasons, skills can issue refinement requests to pull in additional context dynamically.
- Each request triggers a new context assembly cycle with the remaining token budget.
The builtin/context Skill
Actors issue context requests through a context skill that is automatically injected into every actor's skill set:
# Built-in skill, always available skill: name: builtin/context description: "Request additional context during reasoning"anonymous_tools: - name: request_context description: "Request specific context to be added to the conversation" input_schema: type: object properties: query: { type: string, description: "What information do you need?" } focus: { type: array, items: { type: string }, description: "Specific files, classes, or functions to focus on" } breadth: { type: integer, default: 2, description: "How many dependency hops outward (0-5)" } depth: { oneOf: [{ type: integer, minimum: 0 }, { type: string }], default: 3, description: "Detail depth — integer (0-N) or named level from the active DetailLevelMap (e.g., 'SIGNATURES', 'FULL_SOURCE')" } purpose: { type: string, description: "Why do you need this context?" } required: [purpose]
- <span style="color: cyan; font-weight: 600;">name</span>: <span style="color: #66cc66;">query_history</span> <span style="color: cyan; font-weight: 600;">description</span>: <span style="color: #66cc66;">"Query historical context about past decisions and changes"</span> <span style="color: cyan; font-weight: 600;">input_schema</span>: <span style="color: cyan; font-weight: 600;">type</span>: <span style="color: #66cc66;">object</span> <span style="color: cyan; font-weight: 600;">properties</span>: <span style="color: cyan; font-weight: 600;">query</span>: { <span style="color: cyan; font-weight: 600;">type</span>: <span style="color: #66cc66;">string</span> } <span style="color: cyan; font-weight: 600;">scope</span>: { <span style="color: cyan; font-weight: 600;">type</span>: <span style="color: #66cc66;">string</span>, <span style="color: cyan; font-weight: 600;">enum</span>: [<span style="color: #66cc66;">"current_plan"</span>, <span style="color: #66cc66;">"plan_tree"</span>, <span style="color: #66cc66;">"all_plans"</span>], <span style="color: cyan; font-weight: 600;">default</span>: <span style="color: #66cc66;">"plan_tree"</span> } <span style="color: cyan; font-weight: 600;">required</span>: [<span style="color: #66cc66;">query</span>] - <span style="color: cyan; font-weight: 600;">name</span>: <span style="color: #66cc66;">get_context_budget</span> <span style="color: cyan; font-weight: 600;">description</span>: <span style="color: #66cc66;">"Check remaining context token budget"</span> <span style="color: cyan; font-weight: 600;">input_schema</span>: <span style="color: cyan; font-weight: 600;">type</span>: <span style="color: #66cc66;">object</span> <span style="color: cyan; font-weight: 600;">properties</span>: {}
Context Strategy Protocol
Every context strategy implements this protocol. Strategies are pluggable -- built-in strategies ship with CleverAgents, and custom strategies can be registered via configuration.
@runtime_checkable class ContextStrategy(Protocol): """A pluggable context assembly strategy."""<span style="color: cyan;">@property</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">name</span>(<span style="color: cyan;">self</span>) -> <span style="color: cyan;">str</span>: ... <span style="color: cyan;">@property</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">capabilities</span>(<span style="color: cyan;">self</span>) -> <span style="color: cyan;">StrategyCapabilities</span>: ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">can_handle</span>(<span style="color: cyan;">self</span>, request: <span style="color: cyan;">ContextRequest</span>, backends: <span style="color: cyan;">BackendSet</span>) -> <span style="color: cyan;">float</span>: <span style="color: #66cc66;">"""Returns 0.0-1.0 confidence that this strategy can usefully contribute to this request."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">assemble</span>(<span style="color: cyan;">self</span>, request: <span style="color: cyan;">ContextRequest</span>, backends: <span style="color: cyan;">BackendSet</span>, budget: <span style="color: cyan;">int</span>, plan_context: <span style="color: cyan;">PlanContext</span>) -> <span style="color: cyan;">list</span>[<span style="color: cyan;">ContextFragment</span>]: <span style="color: #66cc66;">"""Execute the strategy. Must respect the budget."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">explain</span>(<span style="color: cyan;">self</span>) -> <span style="color: cyan;">str</span>: ...
@dataclass
class StrategyCapabilities:
"""Declares what a strategy is capable of."""
uses_text: bool = False
uses_vector: bool = False
uses_graph: bool = False
uses_temporal: bool = False
uko_levels: list[str] = field(default_factory=list)
resource_types: list[str] = field(default_factory=list)
supports_depth_breadth: bool = False
supports_plan_hierarchy: bool = False
supports_temporal: bool = False
quality_score: float = 0.5
Built-in Strategies
| Strategy | Quality | Backends Required | Description |
|---|---|---|---|
simple-keyword |
0.3 | Text only | Basic keyword/regex text search. Universal fallback. |
semantic-embedding |
0.6 | Vector | Vector similarity search for semantically related content. |
breadth-depth-navigator |
0.85 | Graph | Graph-aware UKO traversal with depth/breadth projection. Primary strategy for code-aware context. |
arce |
0.95 | All | Multi-modal pipeline combining text, vector, and graph with intent analysis. Highest quality. |
temporal-archaeology |
0.5 | Graph + Cold tier | Historical pattern discovery from past decisions and archived context. |
plan-decision-context |
0.7 | Warm/Cold tiers | Retrieves context from parent/ancestor plan decisions. |
Strategy Registration
Strategies are registered through configuration. The context assembly pipeline itself is also fully pluggable — see Architecture > ACMS > Context Assembly Pipeline for the ten pluggable pipeline components.
[context.strategies] enabled = ["simple-keyword", "semantic-embedding", "arce", "breadth-depth-navigator"][context.strategies.custom] "my-domain-strategy" = "my_package.strategies.DomainStrategy"
# Pipeline component overrides (see Architecture > ACMS > Context Assembly Pipeline) [context.pipeline] strategy-selector = "builtin:ConfidenceWeightedSelector" # default budget-allocator = "builtin:ProportionalBudgetAllocator" # default fragment-scorer = "my_extensions.scorers:DomainAwareScorer" # custom override
Hierarchical Plan Context
In a plan hierarchy, each level sees a different slice of the project's resources. Context inheritance flows parent-to-child with progressive focusing. This applies universally across all UKO domains — the same breadth/depth mechanics work for codebases, documents, databases, and infrastructure.
Source Code Example:
| Level | Breadth | Depth | Example Content |
|---|---|---|---|
| Root | Entire project | 0 (MODULE_LISTING) |
Module names, dependency graph edges |
| Subplan A | src/auth/ module + direct deps |
4 (SIGNATURES) |
Function signatures with types, class inheritance chains |
| Sub-subplan A1 | AuthManager class + 2-hop deps |
9 (FULL_SOURCE) |
Complete function bodies, all parameters, all call sites |
Document Example (e.g., a technical specification):
| Level | Breadth | Depth | Example Content |
|---|---|---|---|
| Root | Entire document | 1 (TABLE_OF_CONTENTS_L1) |
Document title + top-level chapter headings |
| Subplan A | "Architecture" chapter + sections that discussesTopic it |
5 (SECTION_OVERVIEWS) |
Section headings + topic keywords + cross-reference targets + one-sentence abstracts |
| Sub-subplan A1 | "ACMS Architecture" section + 2-hop topic references | 10 (FULL_CONTENT) |
Complete section text with all paragraphs, figures, code blocks |
Database Example (e.g., schema migration planning):
| Level | Breadth | Depth | Example Content |
|---|---|---|---|
| Root | Entire database | 1 (TABLE_LISTING) |
Schema names + table and view names |
| Subplan A | auth schema + tables with foreign keys into it |
5 (RELATIONSHIPS) |
Table names + typed columns + constraints + foreign key graph |
| Sub-subplan A1 | users table + dependent views and stored procedures |
9 (FULL_PROCEDURES) |
Complete DDL + triggers + stored procedure bodies + sample data |
The skeleton is a compact, low-depth representation (typically depth 0-1) of the parent's context window that is passed to every child. This ensures children always have the "big picture" available, even though they focus on a narrow area. The skeleton typically consumes 5-15% of the child's token budget (controlled by skeleton_ratio). For a source code resource, the skeleton is a MODULE_LISTING (depth 0) — just module names. For a document resource, the skeleton is a TABLE_OF_CONTENTS_L1 (depth 1) — the table of contents. For a database resource, it is a TABLE_LISTING (depth 1) — schema and table names.
Depth/Breadth Projection
The depth/breadth projection system allows fine-grained control over how much of the knowledge graph is materialized into context. It operates on the UKO graph structure.
- Breadth (integer, 0-N): How many hops in the
uko:references/uko:dependsOn/uko:contains^-1graph to traverse from the focus items. - Depth (integer or named level): How much content to include for each item found. Specified as a raw integer or a named level resolved via the active DetailLevelMap.
- Depth gradient: When enabled, detail decreases with distance from focus.
Source Code Projection Example:
Request: focus=["class://AuthManager"], breadth=2, depth=9 (FULL_SOURCE), gradient=TrueResult (token costs approximate):
Distance 0 — depth 9 (FULL_SOURCE): class AuthManager: # 800 tokens """Manages user authentication...""" def authenticate(self, username, password): ...full body... def validate_token(self, token): ...full body...
Distance 1 — depth 4 (SIGNATURES): class BaseManager: # 120 tokens def connect(self) -> Connection: ... def disconnect(self) -> None: ... class CryptoUtils: # 80 tokens def hash_password(pwd: str) -> str: ... def verify_hash(pwd: str, hash: str) -> bool: ... class UserDB: # 100 tokens def find_user(username: str) -> User | None: ... def create_user(user: User) -> None: ...
Distance 2 — depth 0 (MODULE_LISTING): class Connection # 20 tokens class User # 15 tokens class DatabasePool # 15 tokens
Total: ~1,150 tokens (vs ~15,000 if everything were depth 9)
Document Projection Example:
Request: focus=["uko-doc:section/security-architecture"], breadth=2, depth=10 (FULL_CONTENT), gradient=TrueResult (token costs approximate):
Distance 0 — depth 10 (FULL_CONTENT): ## 5. Security Architecture # 2,400 tokens Complete section text including all paragraphs, code examples, diagrams, and subsections: 5.1 Authentication Flow 5.2 Authorization Model 5.3 Token Management
Distance 1 — depth 6 (TOPIC_SENTENCES — sections that discuss security topics): ## 3. API Design # 180 tokens Section headings + first sentence of each paragraph: "The API uses JWT tokens for authentication..." "Rate limiting is enforced per-client..." ## 8. Deployment # 120 tokens "TLS termination occurs at the load balancer..." "Secrets are injected via Vault..."
Distance 2 — depth 0 (TITLE_ONLY — sections referenced by distance-1 sections): ## 2. System Overview # 30 tokens ## 9. Monitoring # 25 tokens ## Appendix A: Threat Model # 20 tokens
Total: ~2,775 tokens (vs ~18,000 if the entire document were depth 10)
Database Projection Example:
Request: focus=["uko-data:table/auth.users"], breadth=2, depth=11 (FULL_CATALOG), gradient=TrueResult (token costs approximate):
Distance 0 — depth 11 (FULL_CATALOG): CREATE TABLE auth.users ( # 650 tokens id UUID PRIMARY KEY DEFAULT gen_random_uuid(), email VARCHAR(255) UNIQUE NOT NULL, password_hash VARCHAR(60) NOT NULL, ...complete DDL + triggers + sample rows + statistics... );
Distance 1 — depth 7 (DDL — tables with foreign keys to/from users): CREATE TABLE auth.sessions ( # 180 tokens id UUID PRIMARY KEY, user_id UUID REFERENCES auth.users(id), ...DDL with constraints... ); CREATE TABLE auth.roles ( # 120 tokens ...DDL with constraints... ); CREATE VIEW auth.active_users AS ... # 90 tokens
Distance 2 — depth 1 (TABLE_LISTING — tables referenced by distance-1 tables): auth.permissions # 20 tokens auth.audit_log # 15 tokens public.organizations # 15 tokens
Total: ~1,090 tokens (vs ~8,500 if the entire schema were depth 11)
Integration with Plan Lifecycle
Plan Lifecycle ACMS Actions ───────────────── ──────────────────────────── Plan created (agents plan use) ResourceScope resolved. ScopedBackendViews created.Strategize phase begins InitialContextAssembler runs: - Inherits parent context (if subplan) - Runs Context Assembly Pipeline - Produces AssembledContext - Injects into actor's system prompt
Strategy actor reasons Actor may issue ContextRequests via builtin/context skill. Each request triggers re-assembly with remaining budget.
Strategy actor produces decisions Each Decision's context_snapshot captures AssembledContext hash + provenance map.
Execute phase begins New InitialContextAssembler run with execute-phase view. Decisions from Strategize are in warm tier.
Execution actor works Dynamic ContextRequests as needed.
Subplan spawned PlanContextInheritance computes child context from parent. SkeletonCompressor propagates parent context as skeleton. New ResourceScope (possibly narrower).
Apply phase Minimal context assembly (validation results, diff summary).
Plan completes Hot context archived to warm. Warm context ages to cold based on retention policy.
Output Rendering Framework
!!! adr "Architecture Decision" The output rendering framework, format system, and reactive output sessions are defined in ADR-021: CLI and Output Rendering.
Overview
CleverAgents uses a unified Output Rendering Framework to decouple command output data from its visual presentation. Every CLI command produces structured output through a common abstraction layer, and the active format determines how that output is rendered to the terminal (or piped to external consumers). The format is set via the global --format flag, the format config key, or defaults to rich.
The framework is reactive-first: commands do not build a static data structure and hand it to a renderer. Instead, commands open an output session, create element handles for each piece of output (a panel, a table, a progress indicator), and write data to those handles — potentially from multiple concurrent producers. The session coordinates with a materialization strategy selected by the active format, which decides when and how each element's content reaches the terminal. A rich session renders updates in-place as they arrive; a plain session buffers each element and flushes sequentially; a json session accumulates everything and serializes once at the end. Producer code is format-agnostic — it writes to handles without knowing which format is active.
This architecture is designed for modularity, extensibility, and future-proofing — the same session-based output can be consumed by the CLI, a future TUI, a web frontend, or programmatic integrations. The design uses a pipeline of composable stages: session lifecycle management, typed element handles, event-driven materialization, and format-specific element rendering.
Architecture
Rendering Pipeline
All CLI output flows through a five-stage reactive pipeline:
flowchart LR
A["Command Logic"] --> B["OutputSession\n(lifecycle)"]
B --> C["ElementHandles\n(typed producers)"]
C --> D["MaterializationStrategy\n(format-driven policy)"]
D --> E["Terminal / Pipe\n(stdout/stderr)"]
-
Command Logic opens an
OutputSessionand creates typed element handles —PanelHandle,TableHandle,ProgressHandle, etc. — for each piece of output the command will produce. Handles are created in declaration order, which determines the canonical order in which elements appear in sequential formats. -
OutputSession is the central coordinator. It owns the set of active handles, tracks their lifecycle (open → writing → closed), emits
ElementEventobjects to the active materialization strategy, and provides asnapshot()method that returns a staticStructuredOutputrepresenting the accumulated state at any point in time. -
ElementHandles are the producer-facing API. Each handle is typed for a specific element kind (panel, table, tree, etc.) and exposes write methods appropriate to that kind (
add_row(),set_entry(),set_step_status(), etc.). Handles are thread-safe — multiple concurrent coroutines or threads can write to different handles simultaneously. Handles are format-agnostic — the producer never knows or cares which format is active. -
MaterializationStrategy is a polymorphic observer selected by the active format. It receives
ElementEventnotifications from the session and decides when and how to render content. Each strategy delegates the actual visual rendering of an element's accumulated state to a paired ElementRenderer. -
Terminal/Pipe receives the final byte stream. The framework auto-detects whether stdout is a TTY and degrades gracefully (e.g.,
richfalls back totablewhen piped to a non-TTY unless--format richwas explicitly set).
OutputSession
The OutputSession is the core abstraction that replaces direct construction of static output objects. Commands receive a session (typically injected by the CLI framework) and interact with it throughout their execution:
class OutputSession: """A live output document that coordinates element production and materialization. The session manages the lifecycle of all output elements for a single command invocation. It is the bridge between format-agnostic producer code and the format-specific materialization strategy. Thread Safety: The session is thread-safe. Multiple producers may create and write to handles concurrently. The session serializes event delivery to the materialization strategy using an internal event queue. Lifecycle: session = OutputSession.open(command, strategy) handle_a = session.panel("Title") # create handles handle_b = session.table("Results", ...) handle_a.set_entry(...) # write to handles (concurrent OK) handle_b.add_row(...) handle_a.close() # close handles when done handle_b.close() session.close() # finalize the session """<span style="opacity: 0.7;"># --- Session lifecycle ---</span> command: <span style="color: cyan;">str</span> <span style="opacity: 0.7;"># The command that owns this session</span> session_id: <span style="color: cyan;">str</span> <span style="opacity: 0.7;"># Unique session identifier (ULID)</span> created_at: <span style="color: cyan;">datetime</span> <span style="opacity: 0.7;"># Session creation timestamp</span> _strategy: MaterializationStrategy <span style="opacity: 0.7;"># The active materialization strategy</span> _handles: <span style="color: cyan;">OrderedDict</span>[<span style="color: cyan;">str</span>, ElementHandle] <span style="opacity: 0.7;"># handle_id → handle, in declaration order</span> _event_queue: asyncio.Queue[ElementEvent] <span style="opacity: 0.7;"># Internal event queue for serialization</span> _state: SessionState <span style="opacity: 0.7;"># "open" | "closing" | "closed"</span> _lock: threading.Lock <span style="opacity: 0.7;"># Protects handle creation/removal</span> <span style="color: yellow;">@classmethod</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">open</span>(cls, command: <span style="color: cyan;">str</span>, strategy: MaterializationStrategy, metadata: <span style="color: cyan;">dict</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>) -> <span style="color: #66cc66;">"OutputSession"</span>: <span style="color: #66cc66;">"""Open a new output session.</span>Called by the CLI framework before command execution. The strategy is selected based on the resolved format (see Format Resolution). Args: command: The command string (e.g., "project show"). strategy: The materialization strategy for the active format. metadata: Optional command metadata (user, timestamp, etc.). Returns: A new OutputSession ready for element creation. """ ...
<span style="opacity: 0.7;"># --- Element handle factories ---</span> <span style="opacity: 0.7;"># Each factory creates a typed handle, registers it with the session in</span> <span style="opacity: 0.7;"># declaration order, and emits an ElementCreated event to the strategy.</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">panel</span>(self, title: <span style="color: cyan;">str</span>, *, border_style: <span style="color: cyan;">str</span> = <span style="color: #66cc66;">"rounded"</span>, priority: <span style="color: cyan;">str</span> = <span style="color: #66cc66;">"normal"</span>, collapse_hint: <span style="color: cyan;">str</span> = <span style="color: #66cc66;">"auto"</span>, metadata: <span style="color: cyan;">dict</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>) -> <span style="color: #66cc66;">"PanelHandle"</span>: <span style="color: #66cc66;">"""Create a panel element handle for key-value pair output."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">table</span>(self, title: <span style="color: cyan;">str</span> | <span style="color: magenta; font-weight: 600;">None</span>, *, columns: <span style="color: cyan;">list</span>[<span style="color: #66cc66;">"ColumnDef"</span>], summary: <span style="color: cyan;">dict</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>, max_rows_hint: <span style="color: cyan;">int</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>, sort_key: <span style="color: cyan;">str</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>, priority: <span style="color: cyan;">str</span> = <span style="color: #66cc66;">"normal"</span>, collapse_hint: <span style="color: cyan;">str</span> = <span style="color: #66cc66;">"auto"</span>, metadata: <span style="color: cyan;">dict</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>) -> <span style="color: #66cc66;">"TableHandle"</span>: <span style="color: #66cc66;">"""Create a table element handle for tabular data."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">tree</span>(self, root_label: <span style="color: cyan;">str</span>, *, root_style: <span style="color: cyan;">str</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>, max_depth_hint: <span style="color: cyan;">int</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>, show_guides: <span style="color: cyan;">bool</span> = <span style="color: magenta; font-weight: 600;">True</span>, priority: <span style="color: cyan;">str</span> = <span style="color: #66cc66;">"normal"</span>, collapse_hint: <span style="color: cyan;">str</span> = <span style="color: #66cc66;">"auto"</span>, metadata: <span style="color: cyan;">dict</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>) -> <span style="color: #66cc66;">"TreeHandle"</span>: <span style="color: #66cc66;">"""Create a tree element handle for hierarchical data."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">progress</span>(self, label: <span style="color: cyan;">str</span>, *, total: <span style="color: cyan;">int</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>, indeterminate: <span style="color: cyan;">bool</span> = <span style="color: magenta; font-weight: 600;">False</span>, steps: <span style="color: cyan;">list</span>[<span style="color: cyan;">str</span>] | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>, priority: <span style="color: cyan;">str</span> = <span style="color: #66cc66;">"normal"</span>, metadata: <span style="color: cyan;">dict</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>) -> <span style="color: #66cc66;">"ProgressHandle"</span>: <span style="color: #66cc66;">"""Create a progress indicator handle.</span>Args: label: Display label for the progress indicator. total: Total units of work (None for indeterminate). indeterminate: If True, show a spinner instead of a progress bar. steps: Named steps to track (creates ProgressStep objects with initial status "pending"). """ ...
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">status</span>(self, message: <span style="color: cyan;">str</span>, *, level: <span style="color: cyan;">str</span> = <span style="color: #66cc66;">"info"</span>, detail: <span style="color: cyan;">str</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>, priority: <span style="color: cyan;">str</span> = <span style="color: #66cc66;">"normal"</span>, metadata: <span style="color: cyan;">dict</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>) -> <span style="color: #66cc66;">"StatusHandle"</span>: <span style="color: #66cc66;">"""Create a status message handle."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">text</span>(self, content: <span style="color: cyan;">str</span> = <span style="color: #66cc66;">""</span>, *, wrap: <span style="color: cyan;">bool</span> = <span style="color: magenta; font-weight: 600;">True</span>, indent: <span style="color: cyan;">int</span> = 0, priority: <span style="color: cyan;">str</span> = <span style="color: #66cc66;">"normal"</span>, metadata: <span style="color: cyan;">dict</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>) -> <span style="color: #66cc66;">"TextHandle"</span>: <span style="color: #66cc66;">"""Create a text block handle."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">code</span>(self, content: <span style="color: cyan;">str</span> = <span style="color: #66cc66;">""</span>, *, language: <span style="color: cyan;">str</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>, line_numbers: <span style="color: cyan;">bool</span> = <span style="color: magenta; font-weight: 600;">False</span>, highlight_lines: <span style="color: cyan;">list</span>[<span style="color: cyan;">int</span>] | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>, priority: <span style="color: cyan;">str</span> = <span style="color: #66cc66;">"normal"</span>, metadata: <span style="color: cyan;">dict</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>) -> <span style="color: #66cc66;">"CodeHandle"</span>: <span style="color: #66cc66;">"""Create a code block handle."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">diff</span>(self, *, file_a: <span style="color: cyan;">str</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>, file_b: <span style="color: cyan;">str</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>, priority: <span style="color: cyan;">str</span> = <span style="color: #66cc66;">"normal"</span>, metadata: <span style="color: cyan;">dict</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>) -> <span style="color: #66cc66;">"DiffHandle"</span>: <span style="color: #66cc66;">"""Create a diff block handle."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">separator</span>(self, style: <span style="color: cyan;">str</span> = <span style="color: #66cc66;">"line"</span>) -> <span style="color: #66cc66;">"SeparatorHandle"</span>: <span style="color: #66cc66;">"""Create a visual separator. Separators are auto-closed on creation."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">action_hint</span>(self, commands: <span style="color: cyan;">list</span>[<span style="color: cyan;">str</span>], description: <span style="color: cyan;">str</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>) -> <span style="color: #66cc66;">"ActionHintHandle"</span>: <span style="color: #66cc66;">"""Create an action hint. Action hints are auto-closed on creation."""</span> ... <span style="opacity: 0.7;"># --- Session operations ---</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">snapshot</span>(self) -> <span style="color: #66cc66;">"StructuredOutput"</span>: <span style="color: #66cc66;">"""Return a static snapshot of all elements accumulated so far.</span>The snapshot captures the current state of every handle (open or closed) as a StructuredOutput object. This is used by accumulate-mode strategies (json/yaml) at session end, and is available at any time for logging, debugging, or programmatic inspection. """ ...
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">close</span>(self, *, exit_code: <span style="color: cyan;">int</span> = 0) -> <span style="color: #66cc66;">"StructuredOutput"</span>: <span style="color: #66cc66;">"""Close the session and finalize all output.</span>Any handles still open are force-closed (with a warning logged). Emits a SessionEnd event to the strategy. Returns the final StructuredOutput snapshot. Args: exit_code: The command's exit code (included in the snapshot). Returns: The final StructuredOutput snapshot. """ ...
<span style="opacity: 0.7;"># --- Context manager support ---</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">__enter__</span>(self) -> <span style="color: #66cc66;">"OutputSession"</span>: <span style="color: magenta; font-weight: 600;">return</span> self <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">__exit__</span>(self, exc_type, exc_val, exc_tb) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Auto-close session on context exit.</span>
If exiting due to an exception, sets exit_code to 1 and emits an error status element before closing. """ ...
Element Handles
Element handles are the producer-facing API. Each handle type wraps a specific element kind and provides methods appropriate to that kind. All handles share a common base:
class ElementHandle(Generic[E]): """Base class for all element handles. An element handle is a write-only view of an output element. Producers use handles to incrementally build element content without knowledge of the active format or materialization strategy. Type Parameter: E: The OutputElement subclass this handle wraps (e.g., Panel, Table). Thread Safety: Individual handle methods are thread-safe. Multiple threads may call methods on *different* handles concurrently. Concurrent writes to the *same* handle are serialized via an internal lock. Lifecycle: handle = session.table(...) # Created by session factory handle.add_row(...) # Write operations (zero or more) handle.close() # Finalize (required unless auto-closed) Closed Handle Behavior: Calling any write method on a closed handle raises ElementClosedError. """handle_id: <span style="color: cyan;">str</span> <span style="opacity: 0.7;"># Unique handle identifier (ULID)</span> element_type: <span style="color: cyan;">str</span> <span style="opacity: 0.7;"># Semantic type ("panel", "table", etc.)</span> declaration_index: <span style="color: cyan;">int</span> <span style="opacity: 0.7;"># Position in session's declaration order</span> _session: OutputSession <span style="opacity: 0.7;"># Owning session (for event emission)</span> _element: E <span style="opacity: 0.7;"># The accumulated element state</span> _state: HandleState <span style="opacity: 0.7;"># "open" | "closed"</span> _lock: threading.Lock <span style="opacity: 0.7;"># Serializes writes to this handle</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">close</span>(self) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Close this handle, signaling that no more data will be written.</span>Emits an ElementClosed event to the materialization strategy. For buffered strategies, this triggers rendering of the element. """ ...
<span style="color: yellow;">@property</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">is_open</span>(self) -> <span style="color: cyan;">bool</span>: <span style="color: #66cc66;">"""Whether this handle is still accepting writes."""</span> ... <span style="color: yellow;">@property</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">element</span>(self) -> E: <span style="color: #66cc66;">"""The accumulated element state (read-only snapshot)."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">__enter__</span>(self) -> <span style="color: cyan;">Self</span>: <span style="color: magenta; font-weight: 600;">return</span> self <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">__exit__</span>(self, exc_type, exc_val, exc_tb) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Auto-close handle on context exit."""</span> <span style="color: magenta; font-weight: 600;">if</span> self.is_open: self.close()class PanelHandle(ElementHandle[Panel]): """Handle for building a Panel element incrementally. Panels are titled groups of key-value pairs. Entries can be added, updated, or removed after creation. """
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">set_entry</span>(self, key: <span style="color: cyan;">str</span>, value: <span style="color: cyan;">str</span>, *, style_hint: <span style="color: cyan;">str</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>, icon: <span style="color: cyan;">str</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Set or update a key-value entry in the panel.</span>If an entry with the given key already exists, it is updated. Otherwise, a new entry is appended. Emits an ElementUpdated event. """ ...
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">set_entries</span>(self, entries: <span style="color: cyan;">dict</span>[<span style="color: cyan;">str</span>, <span style="color: cyan;">str</span>], *, style_hints: <span style="color: cyan;">dict</span>[<span style="color: cyan;">str</span>, <span style="color: cyan;">str</span>] | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Set multiple entries at once (batch update). Emits a single event."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">remove_entry</span>(self, key: <span style="color: cyan;">str</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Remove an entry by key. Emits an ElementUpdated event."""</span> ...class TableHandle(ElementHandle[Table]): """Handle for building a Table element incrementally. Tables are the primary element for streamed data. Rows can be added one at a time or in batches as data becomes available from queries, API calls, or concurrent operations. """
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">add_row</span>(self, row: <span style="color: cyan;">dict</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Append a single row to the table.</span>The row dict maps column names to cell values. Missing columns are filled with None. Extra columns not in the schema are ignored. Emits an ElementUpdated event (type=row_added). """ ...
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">add_rows</span>(self, rows: <span style="color: cyan;">list</span>[<span style="color: cyan;">dict</span>]) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Append multiple rows in a batch. Emits a single ElementUpdated event."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">set_summary</span>(self, summary: <span style="color: cyan;">dict</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Set or update the summary/aggregation row. Emits an ElementUpdated event."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">set_sort_key</span>(self, column: <span style="color: cyan;">str</span>, *, descending: <span style="color: cyan;">bool</span> = <span style="color: magenta; font-weight: 600;">False</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Change the sort key. Emits an ElementUpdated event."""</span> ...class TreeHandle(ElementHandle[Tree]): """Handle for building a Tree element incrementally. Trees are built by adding child nodes to existing nodes. The root node is created with the handle. Subtrees can be constructed incrementally as hierarchical data is discovered. """
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">add_child</span>(self, parent_path: <span style="color: cyan;">str</span> | <span style="color: magenta; font-weight: 600;">None</span>, label: <span style="color: cyan;">str</span>, *, style_hint: <span style="color: cyan;">str</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>, collapsed: <span style="color: cyan;">bool</span> = <span style="color: magenta; font-weight: 600;">False</span>, metadata: <span style="color: cyan;">dict</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>) -> <span style="color: cyan;">str</span>: <span style="color: #66cc66;">"""Add a child node to the tree.</span>Args: parent_path: Slash-separated path to the parent node (None = root). label: Display label for the new node. style_hint: Optional color/style hint. collapsed: Whether this node starts collapsed in interactive renderers. metadata: Arbitrary data attached to the node. Returns: The full path to the newly created node (for use as parent_path in subsequent add_child calls). Emits an ElementUpdated event. """ ...
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">set_node_style</span>(self, path: <span style="color: cyan;">str</span>, style_hint: <span style="color: cyan;">str</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Update the style of an existing node. Emits an ElementUpdated event."""</span> ...class ProgressHandle(ElementHandle[ProgressIndicator]): """Handle for updating a ProgressIndicator element. Progress handles are unique in that they are expected to receive many rapid updates. The materialization strategy may throttle update events to avoid overwhelming the terminal (e.g., limiting redraws to 10/sec). """
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">set_progress</span>(self, current: <span style="color: cyan;">int</span>, total: <span style="color: cyan;">int</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Update the progress counter.</span>Args: current: Current progress value. total: Total value (can change, e.g., when total is discovered late). Emits an ElementUpdated event (may be throttled by the strategy). """ ...
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">set_step_status</span>(self, step_label: <span style="color: cyan;">str</span>, status: <span style="color: cyan;">str</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Update the status of a named step.</span>Args: step_label: The label of the step to update. status: New status — "pending" | "active" | "done" | "error" | "skipped". Emits an ElementUpdated event. """ ...
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">set_label</span>(self, label: <span style="color: cyan;">str</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Update the progress label text. Emits an ElementUpdated event."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">increment</span>(self, delta: <span style="color: cyan;">int</span> = 1) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Increment progress by delta. Convenience wrapper around set_progress."""</span> ...class StatusHandle(ElementHandle[StatusMessage]): """Handle for a status message. Status handles are typically created and immediately closed (fire-and-forget messages). However, they can be kept open for messages that may be revised (e.g., a "Working..." status that becomes "Done" or "Failed"). """
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">set_message</span>(self, message: <span style="color: cyan;">str</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Update the status message text. Emits an ElementUpdated event."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">set_level</span>(self, level: <span style="color: cyan;">str</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Change the status level. Emits an ElementUpdated event."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">set_detail</span>(self, detail: <span style="color: cyan;">str</span> | <span style="color: magenta; font-weight: 600;">None</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Set or clear the detail text. Emits an ElementUpdated event."""</span> ...class TextHandle(ElementHandle[TextBlock]): """Handle for a text block. Supports appending text incrementally."""
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">append</span>(self, text: <span style="color: cyan;">str</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Append text to the block. Emits an ElementUpdated event."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">set_content</span>(self, content: <span style="color: cyan;">str</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Replace the entire content. Emits an ElementUpdated event."""</span> ...class CodeHandle(ElementHandle[CodeBlock]): """Handle for a code block."""
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">set_content</span>(self, content: <span style="color: cyan;">str</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Set the code content. Emits an ElementUpdated event."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">set_language</span>(self, language: <span style="color: cyan;">str</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Set the language for syntax highlighting. Emits an ElementUpdated event."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">set_highlight_lines</span>(self, lines: <span style="color: cyan;">list</span>[<span style="color: cyan;">int</span>]) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Set lines to highlight. Emits an ElementUpdated event."""</span> ...class DiffHandle(ElementHandle[DiffBlock]): """Handle for a diff block. Hunks can be added incrementally."""
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">add_hunk</span>(self, header: <span style="color: cyan;">str</span>, lines: <span style="color: cyan;">list</span>[<span style="color: #66cc66;">"DiffLine"</span>]) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Add a diff hunk. Emits an ElementUpdated event."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">set_stats</span>(self, insertions: <span style="color: cyan;">int</span>, deletions: <span style="color: cyan;">int</span>, **extra: <span style="color: cyan;">int</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Set diff statistics. Emits an ElementUpdated event."""</span> ...
Element Events
Element handles communicate with the materialization strategy through a typed event system. Events are the sole interface between production (handles) and consumption (strategy) — this indirection is what enables format-agnostic producer code:
class ElementEvent: """Base class for all events emitted by element handles.""" event_type: str # "created" | "updated" | "closed" handle_id: str # The handle that emitted this event element_type: str # The element kind ("panel", "table", etc.) timestamp: datetime # When the event occurred session_id: str # The owning sessionclass ElementCreated(ElementEvent): """Emitted when a new element handle is created via a session factory.""" event_type = "created" declaration_index: int # Position in session declaration order initial_state: OutputElement # The element's initial state
class ElementUpdated(ElementEvent): """Emitted when data is written to an element handle.""" event_type = "updated" update_type: str # Kind-specific: "entry_set", "row_added", # "progress_changed", "step_status_changed", etc. delta: dict # The change payload (what was added/modified) element_snapshot: OutputElement # The full element state after this update
class ElementClosed(ElementEvent): """Emitted when an element handle is closed (no more data will arrive).""" event_type = "closed" final_state: OutputElement # The element's final accumulated state
class SessionEnd(ElementEvent): """Emitted when the session itself is closed.""" event_type = "session_end" exit_code: int snapshot: "StructuredOutput" # The complete accumulated output
Element Data Model (Snapshot Types)
Each element handle accumulates state into a typed snapshot object. These are the data classes that represent a fully-built element — they are what the ElementRenderer receives when it is time to paint. They are also the building blocks of the StructuredOutput returned by session.snapshot():
class OutputElement: """Base class for all output element snapshot types.""" element_type: str # Semantic type identifier metadata: dict # Arbitrary metadata (timestamps, IDs, etc.) priority: str = "normal" # "critical" | "normal" | "supplementary" collapse_hint: str = "auto" # "always" | "auto" | "never" — guidance for renderers # on whether this element can be collapsed/hiddenclass Panel(OutputElement): """A titled group of key-value pairs.""" element_type = "panel" title: str entries: list[PanelEntry] # Each entry: key, value, style_hint (color, icon, etc.) border_style: str = "rounded" # "rounded" | "square" | "heavy" | "none"
class PanelEntry: """A single key-value pair within a Panel.""" key: str value: str style_hint: str | None = None # Color/style for the value (e.g., "success", "warning") icon: str | None = None # Optional icon/prefix character
class Table(OutputElement): """A tabular data set with typed columns.""" element_type = "table" title: str | None columns: list[ColumnDef] # name, type, alignment, width_hint, sortable rows: list[dict] # Column name → cell value summary: dict | None # Optional aggregation row (totals, counts) max_rows_hint: int | None # Suggest truncation for large datasets sort_key: str | None # Default sort column
class ColumnDef: """Schema for a single table column.""" name: str # Column display name type: str = "string" # "string" | "number" | "boolean" | "datetime" | "id" alignment: str = "left" # "left" | "right" | "center" width_hint: int | None = None # Suggested character width (None = auto) sortable: bool = False # Whether this column can be sorted style_hint: str | None = None # Default style for cells in this column
class Tree(OutputElement): """A hierarchical tree structure.""" element_type = "tree" root: TreeNode # Recursive node structure max_depth_hint: int | None # Suggest depth truncation show_guides: bool = True # Whether to show tree guide lines
class TreeNode: """A node in a tree structure.""" label: str style_hint: str | None # Color/style for this node children: list["TreeNode"] collapsed: bool = False # Hint: start collapsed in interactive renderers metadata: dict # Arbitrary data attached to the node
class StatusMessage(OutputElement): """A status line (success, warning, error, info).""" element_type = "status" level: str # "ok" | "warn" | "error" | "info" message: str detail: str | None # Optional detail text
class ProgressIndicator(OutputElement): """A progress bar or spinner for long-running operations.""" element_type = "progress" label: str current: int | None total: int | None indeterminate: bool = False # Spinner mode vs. progress bar mode steps: list[ProgressStep] | None # Named steps with status (pending/active/done)
class ProgressStep: """A named step within a progress indicator.""" label: str status: str # "pending" | "active" | "done" | "error" | "skipped"
class CodeBlock(OutputElement): """A block of source code with optional syntax highlighting.""" element_type = "code" content: str language: str | None # For syntax highlighting line_numbers: bool = False highlight_lines: list[int] | None # Lines to emphasize
class DiffBlock(OutputElement): """A unified diff display.""" element_type = "diff" hunks: list[DiffHunk] file_a: str | None file_b: str | None stats: dict | None # insertions, deletions, etc.
class DiffHunk: """A single hunk within a diff.""" header: str # @@ line range @@ lines: list[DiffLine]
class DiffLine: """A single line in a diff hunk.""" type: str # "context" | "add" | "remove" content: str line_number_old: int | None line_number_new: int | None
class TextBlock(OutputElement): """A free-form text block (descriptions, rationale, etc.).""" element_type = "text" content: str wrap: bool = True indent: int = 0
class Separator(OutputElement): """A visual separator between logical groups.""" element_type = "separator" style: str = "line" # "line" | "blank" | "double"
class ActionHint(OutputElement): """A suggested next-step action for the user.""" element_type = "action_hint" commands: list[str] # Suggested CLI commands description: str | None
class StructuredOutput: """Static snapshot of a complete command output. This is the accumulated state of all elements at a point in time. It is produced by OutputSession.snapshot() and OutputSession.close(). Uses: - Final serialization for json/yaml formats - Logging and audit trails - Programmatic inspection and testing - TUI widget data binding (initial state) """ command: str # The command that produced this output session_id: str # The session that produced this output elements: list[OutputElement] # Ordered list of element snapshots exit_code: int = 0 timing: dict | None # start_time, end_time, duration metadata: dict # command-specific metadata
Materialization Strategies
The materialization strategy is the format-side counterpart to the output session. It receives element events and decides when and how to render content. Each strategy is paired with an ElementRenderer that handles the actual visual formatting of individual elements.
The strategy pattern creates a clean separation between timing/ordering policy (when to render) and visual formatting (how to render). This means the same PlainElementRenderer can be used whether elements arrive all-at-once or are streamed concurrently — the strategy handles the coordination.
class MaterializationStrategy(Protocol): """Interface for format-driven output materialization. A materialization strategy receives element lifecycle events from the OutputSession and decides when to render element content to the output stream. Strategies do not render elements themselves — they delegate to a paired ElementRenderer at the appropriate time. The strategy is the mechanism by which format-agnostic producer code produces correct output regardless of format. The producer writes to handles; the strategy decides what reaches the terminal and when. """strategy_name: <span style="color: cyan;">str</span> <span style="opacity: 0.7;"># "live" | "sequential_buffer" | "accumulate"</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">bind</span>(self, renderer: <span style="color: #66cc66;">"ElementRenderer"</span>, terminal_caps: <span style="color: #66cc66;">"TerminalCapabilities"</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Bind this strategy to a renderer and terminal capabilities.</span>Called once during format resolution, before the session opens. The strategy retains a reference to the renderer for use during event handling. The output stream is provided separately via on_session_begin, since the session owns the stream. """ ...
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">on_session_begin</span>(self, session: OutputSession, stream: <span style="color: cyan;">IO</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Called when the session opens. The strategy receives the output</span>stream and may write preamble (e.g., opening JSON bracket).""" ...
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">on_element_created</span>(self, event: ElementCreated) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Called when a new element handle is created."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">on_element_updated</span>(self, event: ElementUpdated) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Called when data is written to an element handle."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">on_element_closed</span>(self, event: ElementClosed) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Called when an element handle is closed."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">on_session_end</span>(self, event: SessionEnd) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Called when the session closes. The strategy may write epilogue."""</span> ...class LiveMaterializer(MaterializationStrategy): """Materialization strategy for the
richformat. Renders element updates in real-time using terminal cursor movement. Multiple elements can be visually active and updating simultaneously. The terminal display is a live document that is rewritten in place. Behavior: - on_element_created: Allocates screen region for the element, renders initial (possibly empty) visual state. - on_element_updated: Re-renders the element in place using cursor movement. For progress indicators, updates may be throttled to a maximum refresh rate (default: 15 fps) to avoid terminal flooding. - on_element_closed: Renders the final state and freezes the screen region (no further updates). May apply a visual transition (e.g., spinner resolves to a checkmark). - on_session_end: Finalizes the display, moves cursor to end, and restores normal terminal scrolling. Screen Layout: Elements are arranged vertically in declaration order. Each element occupies a contiguous block of terminal lines. The materializer tracks the line offset and height of each element's region. When an element's height changes (e.g., a table gains rows), subsequent elements are shifted down. Concurrent Updates: Updates from multiple handles are coalesced into a single frame refresh at the target frame rate. The materializer maintains a dirty-element set and redraws all dirty elements in a single pass per frame. """ strategy_name = "live"_frame_rate: <span style="color: cyan;">float</span> = 15.0 <span style="opacity: 0.7;"># Maximum redraws per second</span> _element_regions: <span style="color: cyan;">OrderedDict</span>[<span style="color: cyan;">str</span>, ScreenRegion] <span style="opacity: 0.7;"># handle_id → screen region</span> _dirty_set: <span style="color: cyan;">set</span>[<span style="color: cyan;">str</span>] <span style="opacity: 0.7;"># handle_ids that need redraw</span> _frame_timer: asyncio.TimerHandle <span style="opacity: 0.7;"># Coalescing timer for frame redraws</span>class SequentialBufferMaterializer(MaterializationStrategy): """Materialization strategy for
plain,color, andtableformats. Buffers element content and renders elements sequentially in declaration order. An element's content is rendered to the output stream only when its handle is closed. If handles are closed out of declaration order, the out-of-order element's rendered content is held in a buffer until all preceding elements have been rendered. This strategy ensures that static, scrolling output formats produce coherent sequential output even when producers write to handles concurrently and close them in arbitrary order. Behavior: - on_element_created: Records the element's declaration index. No output. - on_element_updated: Buffers the update internally. No output. - on_element_closed: If this element is the next in declaration order, renders it immediately (and any buffered subsequent elements that are also closed). Otherwise, buffers the rendered content. - on_session_end: Force-renders any remaining buffered elements (handles that were never closed, in declaration order). Example (two tables populated concurrently): 1. Handle A (index 0) created — table "Resources" 2. Handle B (index 1) created — table "Validations" 3. Handle B receives rows, Handle A receives rows (interleaved) 4. Handle B closes (index 1) — rendered content buffered (waiting for A) 5. Handle A closes (index 0) — A is rendered to stream, then buffered B is rendered to stream Result: Output shows table A followed by table B, regardless of the order in which data arrived or handles closed. """ strategy_name = "sequential_buffer"_next_render_index: <span style="color: cyan;">int</span> = 0 <span style="opacity: 0.7;"># The declaration index to render next</span> _rendered_buffers: <span style="color: cyan;">dict</span>[<span style="color: cyan;">int</span>, <span style="color: cyan;">str</span>] <span style="opacity: 0.7;"># index → pre-rendered content (waiting)</span> _closed_set: <span style="color: cyan;">set</span>[<span style="color: cyan;">int</span>] <span style="opacity: 0.7;"># Declaration indices of closed elements</span>
class AccumulateMaterializer(MaterializationStrategy): """Materialization strategy forjsonandyamlformats. Accumulates all element data silently until the session ends, then serializes the complete StructuredOutput as a single JSON or YAML document. Behavior: - on_element_created: No output. - on_element_updated: No output. - on_element_closed: No output. - on_session_end: Calls session.snapshot() to get the final StructuredOutput, then delegates to the ElementRenderer's serialize() method for complete document serialization. This strategy is the simplest — it ignores all intermediate events and only acts on session_end. It exists as a distinct strategy (rather than a special case) to maintain the uniform strategy interface. """ strategy_name = "accumulate"
ElementRenderer Protocol
While the MaterializationStrategy controls when elements are rendered, the ElementRenderer controls how each element type is visually formatted. Each format has a paired ElementRenderer implementation:
class ElementRenderer(Protocol): """Interface for format-specific element rendering. An ElementRenderer knows how to paint each element type for a specific output format. It is called by the MaterializationStrategy when it is time to render an element. Implementations: - PlainElementRenderer: ASCII text, no escapes - ColorElementRenderer: ANSI-colored text, same layout as plain - TableElementRenderer: Unicode box-drawing with color - RichElementRenderer: Advanced terminal features (cursor, animation) - JsonElementRenderer: JSON serialization - YamlElementRenderer: YAML serialization """format_name: <span style="color: cyan;">str</span> <span style="opacity: 0.7;"># "plain", "color", "table", "rich", "json", "yaml"</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">render_panel</span>(self, panel: Panel, stream: <span style="color: cyan;">IO</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Render a panel element to the stream."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">render_table</span>(self, table: Table, stream: <span style="color: cyan;">IO</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Render a table element to the stream."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">render_tree</span>(self, tree: Tree, stream: <span style="color: cyan;">IO</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Render a tree element to the stream."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">render_status</span>(self, status: StatusMessage, stream: <span style="color: cyan;">IO</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Render a status message to the stream."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">render_progress</span>(self, progress: ProgressIndicator, stream: <span style="color: cyan;">IO</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Render a progress indicator to the stream."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">render_code</span>(self, code: CodeBlock, stream: <span style="color: cyan;">IO</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Render a code block to the stream."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">render_diff</span>(self, diff: DiffBlock, stream: <span style="color: cyan;">IO</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Render a diff block to the stream."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">render_text</span>(self, text: TextBlock, stream: <span style="color: cyan;">IO</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Render a text block to the stream."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">render_separator</span>(self, separator: Separator, stream: <span style="color: cyan;">IO</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Render a visual separator to the stream."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">render_action_hint</span>(self, hint: ActionHint, stream: <span style="color: cyan;">IO</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Render an action hint to the stream."""</span> ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">render_element</span>(self, element: OutputElement, stream: <span style="color: cyan;">IO</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Dispatch to the appropriate render method based on element type.</span>This is the primary entry point used by materialization strategies. It uses a dispatch table to route to the correct typed method. """ dispatch = { "panel": self.render_panel, "table": self.render_table, "tree": self.render_tree, "status": self.render_status, "progress": self.render_progress, "code": self.render_code, "diff": self.render_diff, "text": self.render_text, "separator": self.render_separator, "action_hint": self.render_action_hint, } handler = dispatch.get(element.element_type) if handler: handler(element, stream)
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">serialize</span>(self, output: StructuredOutput, stream: <span style="color: cyan;">IO</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Serialize a complete StructuredOutput to the stream.</span>Used by AccumulateMaterializer for json/yaml formats. For visual formats (plain/color/table/rich), this method iterates over output.elements and calls render_element for each. """ ...
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">can_render</span>(self, terminal_caps: <span style="color: #66cc66;">"TerminalCapabilities"</span>) -> <span style="color: cyan;">bool</span>: <span style="color: #66cc66;">"""Whether this renderer can operate in the given terminal environment."""</span> ...
Format Resolution
The active format is resolved using a precedence chain:
- CLI flag:
--format <value>on the command line (highest priority). - Environment variable:
CLEVERAGENTS_FORMAT=<value>. - Config file: The
core.formatkey in the global config (agents config set core.format <value>). - TTY detection: If stdout is not a TTY and no explicit format was set, fall back to
plain(notrich), since non-TTY consumers cannot interpret ANSI codes or cursor movement. - Default:
rich.
Once the format is resolved, the CLI framework selects the corresponding (MaterializationStrategy, ElementRenderer) pair from the RendererRegistry, opens an OutputSession bound to that strategy, and passes the session to the command implementation.
Format Specifications
plain — Plain Text
Philosophy: Maximum portability. Output is pure ASCII text with no escape codes, no box-drawing characters, and no color. Suitable for piping to files, logs, grep, awk, or any non-terminal consumer.
Rendering rules:
- Panels: Rendered as indented key-value pairs with a header line.
- Tables: Rendered as aligned columns separated by whitespace (no box drawing). Column headers are separated from data by a dashed line.
- Trees: Rendered with ASCII indentation using
+--and|characters. - Status messages: Prefixed with
[OK],[WARN],[ERROR],[INFO]. - Progress: Rendered as static status lines (no animation). Steps shown as
[x](done),[ ](pending),[>](active). - Diffs: Standard unified diff format.
- Code blocks: Raw text with optional line numbers.
- No ANSI escape codes of any kind.
- No Unicode characters beyond basic ASCII (no box drawing, no checkmarks, no arrows).
Example (agents --format plain project show local/api-service):
$ agents --format plain project show local/api-serviceProject Details Name: local/api-service Description: Backend API Resources: 2 Remote: no Created: 2026-02-08 12:46
Linked Resources Resource Type Sandbox Read-Only
local/api-repo git-checkout git_worktree no local/staging-db local/database transaction_rollback yes
Validations (3) local/run-tests pytest --cov=src --cov-fail-under=80 required local/lint-check ruff check . required local/check-bundle-size node scripts/check-bundle-size.js info
Context Include: repo Exclude: /node_modules/ Max File Size: 1 MB
Indexing Status Text Index: ready Vector Index: ready Graph Store: disabled Indexed Files: 347 Last Indexed: 12:48
Active Plans Plan ID Action Phase
01HXM7A9 local/code-coverage execute
[OK] Project loaded
Example (agents --format plain plan list):
$ agents --format plain plan list --phase executePlans ID Phase State Action Project Elapsed
01HXM7A9 execute processing local/code-coverage local/api-service 00:01:12
Filters Phase: execute State: (any) Project: (any) Action: (any)
Summary Total: 1 Processing: 1 Completed: 0 Errored: 0
[OK] 1 plan listed
color — Colored Plain Text
Philosophy: Same structural layout as plain, but with ANSI color codes applied to improve readability. No box-drawing characters, no cursor movement, no animation.
Rendering rules:
- Identical layout to
plain, but with color applied:- Headers/titles: Bold cyan.
- Keys: Bold blue.
- Values: Default color, with semantic coloring:
- Success/positive: Green.
- Warnings/attention: Yellow.
- Errors/failures: Red.
- Identifiers/names: Cyan.
- Counts/numbers: Default (white).
- Table headers: Bold cyan with underlines rendered as dim dashes.
- Status prefixes:
[OK]in green,[WARN]in yellow,[ERROR]in red,[INFO]in blue. - Diff lines:
+lines green,-lines red,@@headers cyan.
- No box-drawing characters — uses the same whitespace/dash layout as
plain. - No cursor movement or animation — pure scrolling output.
- Respects
NO_COLORenvironment variable: IfNO_COLORis set,colorformat falls back toplain.
Example (agents --format color project show local/api-service):
$ agents --format color project show local/api-serviceProject Details Name: local/api-service Description: Backend API Resources: 2 Remote: no Created: 2026-02-08 12:46
Linked Resources Resource Type Sandbox Read-Only ---------------- -------------- -------------------- --------- local/api-repo git-checkout git_worktree no local/staging-db local/database transaction_rollback yes
Validations (3) local/run-tests pytest --cov=src --cov-fail-under=80 required local/lint-check ruff check . required local/check-bundle-size node scripts/check-bundle-size.js info
Context Include: repo Exclude: /node_modules/ Max File Size: 1 MB
Indexing Status Text Index: ready Vector Index: ready Graph Store: disabled Indexed Files: 347 Last Indexed: 12:48
Active Plans Plan ID Action Phase -------- ------------------- ------- 01HXM7A9 local/code-coverage execute
[OK] Project loaded
table — ASCII Box-Drawing Tables
Philosophy: Structured, visually distinct panels and tables using Unicode box-drawing characters (╭──────────────────╮╰╯│─). Uses color. This is the style shown in the existing CLI examples throughout this document.
Rendering rules:
- Panels: Rendered as bordered boxes with title in the top border. Uses
╭─╮│╰──────────────────╯characters for rounded corners. - Tables: Rendered inside bordered boxes with column-aligned headers and separator lines using
─and│. - Trees: Rendered inside a bordered box using
├──,└──,│tree guide characters. - Status messages: Use Unicode indicators:
✓(green) for OK,⚠(yellow) for WARN,✗(red) for ERROR,ℹ(blue) for INFO. - Progress: Rendered as a step list inside a panel with
✓,⏳,•markers. - Color scheme: Same semantic coloring as
colorformat, applied within box structures. - No animation or cursor movement — the boxes are static, scrolling output.
Distinction from rich: The table format uses the same box-drawing panels and color as rich for static content, but it does not use any dynamic or interactive terminal features. There are no animated spinners, no live-updating progress bars, no cursor movement, and no in-place redraws. All output is static and scrolls sequentially. Where rich would show a spinning ⠋ and a live progress bar, table renders a static snapshot using fixed markers (✓, ⏳, •). This makes table suitable for terminals without advanced capabilities, and for output that will be reviewed after the fact (e.g., scrollback buffers).
Example (agents --format table project show local/api-service):
$ agents --format table project show local/api-service╭─ Project Details ──────────────╮ │ Name: local/api-service │ │ Description: Backend API │ │ Resources: 2 │ │ Remote: no │ │ Created: 2026-02-08 12:46 │ ╰────────────────────────────────╯
╭─ Linked Resources ──────────────────────────────────────────────────────╮ │ Resource Type Sandbox Read-Only │ │ ──────────────── ────────────── ──────────────────── ───────── │ │ local/api-repo git-checkout git_worktree no │ │ local/staging-db local/database transaction_rollback yes │ ╰─────────────────────────────────────────────────────────────────────────╯
╭─ Validations (3) ────────────────────────────────────────────────────╮ │ local/run-tests pytest --cov=src --cov-fail-under=80 required │ │ local/lint-check ruff check . required │ │ local/check-bundle-size node scripts/check-bundle-size.js info │ ╰──────────────────────────────────────────────────────────────────────╯
╭─ Context ───────────────────╮ │ Include: repo │ │ Exclude: /node_modules/ │ │ Max File Size: 1 MB │ ╰─────────────────────────────╯
╭─ Indexing Status ──────────╮ │ Text Index: ready │ │ Vector Index: ready │ │ Graph Store: disabled │ │ Indexed Files: 347 │ │ Last Indexed: 12:48 │ ╰────────────────────────────╯
╭─ Active Plans ──────────────────────────╮ │ Plan ID Action Phase │ │ ──────── ─────────────────── ─────── │ │ 01HXM7A9 local/code-coverage execute │ ╰─────────────────────────────────────────╯
✓ OK Project loaded
Example (agents --format table plan execute 01HXM8C2ZK):
Unlike rich mode which would show animated spinners and a live progress bar, the table format renders a static snapshot of execution state:
$ agents --format table plan execute 01HXM8C2ZK╭─ Execution ──────────────────────╮ │ Plan: 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ │ Phase: execute │ │ Sandbox: git_worktree │ │ Worker: local/executor │ │ Started: 12:58:10 │ │ Attempt: 1 │ ╰──────────────────────────────────╯
╭─ Strategy Summary ─────────────────────╮ │ Decisions: 8 │ │ Invariants: 2 │ │ Planned Child Plans: 2+ │ │ Estimated Files: ~12 │ │ Risk: low │ ╰────────────────────────────────────────╯
╭─ Progress ────────╮ │ ✓ Collect context │ │ ✓ Run tools │ │ ⏳ Build changeset │ │ • Validate │ ╰───────────────────╯
✓ OK Execution started
rich — Modern Rich CLI Elements
Philosophy: The premium interactive terminal experience. Uses advanced terminal capabilities: cursor movement, inline updates, animated spinners, live-updating progress bars, collapsible sections, syntax highlighting, and dynamic layout. This is the default format.
Rendering rules:
- Panels: Rich bordered panels with rounded corners, title bars, and optional collapse/expand behavior. Panels may animate into view.
- Tables: Full-featured tables with automatic column sizing, truncation with ellipsis, sortable column indicators, alternating row shading, and horizontal scrolling for wide tables.
- Trees: Interactive collapsible trees. Nodes expand/collapse with visual animation. Color-coded by node type. Depth guides use dotted lines.
- Status messages: Use animated checkmarks/spinners that resolve to final state. Success messages may briefly flash or highlight.
- Progress: Live-updating progress bars with:
- Animated spinners (Braille, dots, or bars depending on terminal capability).
- Elapsed time and ETA.
- Per-step status with animated transitions (pending → active → done).
- Multi-line progress for parallel operations.
- Diffs: Syntax-highlighted side-by-side or unified diffs with line numbers, change highlighting at the character level (not just line level), and navigable hunks.
- Code blocks: Full syntax highlighting using terminal colors (256-color or truecolor when available). Line numbers in dim color. Highlighted lines with background color.
- Dynamic layout: Adapts to terminal width. Narrow terminals get a stacked layout; wide terminals get side-by-side panels.
- Live updates: Long-running commands (plan execute, plan status) use live-updating displays that redraw in place rather than scrolling.
- Graceful degradation: If the terminal does not support required capabilities (e.g., no truecolor, no cursor movement), the renderer automatically falls back to
tablerendering for those elements.
Example (agents --format rich plan execute 01HXM8C2ZK):
The rich format produces output that cannot be fully represented in static documentation — animated spinners cycle in place, progress bars fill smoothly, and elements update without scrolling. The rendering below is a static snapshot of what the terminal would display at a given moment:
$ agents --format rich plan execute 01HXM8C2ZK╭─ Execution ──────────────────────╮ │ Plan: 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ │ Phase: execute │ │ Sandbox: git_worktree │ │ Worker: local/executor │ │ Started: 12:58:10 │ ╰──────────────────────────────────╯
⠋ Collecting context... (⠋⠙⠹⠸⠼⠴⠦⠧⠇⠏ animates in place) ├── repo: local/api-repo ✓ └── db: local/staging-db ✓
╭─ Strategy Summary ──────────────────────────────────────────────────────╮ │ 8 decisions │ 2 invariants │ 2+ child plans │ ~12 files │ risk: low │ ╰─────────────────────────────────────────────────────────────────────────╯
Progress ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 42% elapsed 0:01:12 ETA 0:01:40 ✓ Collect context .................. 0.8s ✓ Run tools (8 calls) ............. 12.4s ⠙ Build changeset ................. (running) (animates in place) ○ Validate ........................ (pending)
╭─ Live Tool Calls ───────────────────────────────────────╮ │ #6 read_file src/auth/__init__.py ✓ 0.1s │ │ #7 write_file tests/test_auth.py ✓ 0.2s │ │ #8 edit_file src/auth/session.py ⠙ ... │ ╰─────────────────────────────────────────────────────────╯
In rich mode:
- The Braille spinner characters (
⠋⠙⠹⠸⠼⠴⠦⠧⠇⠏) animate in real-time, cycling in place without scrolling. - The progress bar (
━━━) fills smoothly as work completes. - Completed steps appear with a green
✓via in-place line update (the line rewrites, it does not scroll). - The "Live Tool Calls" panel scrolls its content internally, showing only the most recent N calls.
- The terminal is not flooded with scrolling text — elements update in place using cursor movement.
Example (agents --format rich version):
$ agents --format rich version╭─────────────────────────────────────╮ │ CleverAgents CLI v1.0.0 │ │ channel: stable │ ╰─────────────────────────────────────╯
╭─ Build ─────────────────────────────╮ │ Build Date: 2026-02-08 │ │ Commit: a17c3f9 │ │ Schema: v3 │ │ Platform: linux-x86_64 │ │ Python: 3.13.1 │ ╰─────────────────────────────────────╯
╭─ Dependencies ─────────────────────────────────────────────────────────╮ │ LangGraph 0.2.60 │ LangChain 0.3.18 │ MCP SDK 1.4.0 │ Pydantic 2.10.4 │ ╰────────────────────────────────────────────────────────────────────────╯
✓ OK Version reported
In rich mode, the version card may use background colors, bold gradients, or subtle box shadows (depending on terminal truecolor support). Elements may animate into view with a brief slide or fade transition.
json — JSON Data Structure
Philosophy: Machine-readable output for programmatic consumption. Every command produces a well-defined JSON object. No ANSI color codes in the structural data. Color codes may appear only within text values that represent verbatim content (such as code blocks or diff output where the original text contained ANSI sequences), but all structural keys, labels, and metadata are plain strings.
Rendering rules:
- Top-level structure: Always a JSON object with a standard envelope:
{ "command": "project show", "status": "ok", "exit_code": 0, "data": { ... }, "timing": { "duration_ms": 42 }, "metadata": { ... } } - Panels: Rendered as nested objects within
data. - Tables: Rendered as arrays of objects within
data. - Trees: Rendered as nested objects with
childrenarrays. - Status messages: Included in a
messagesarray within the envelope. - Progress: Not rendered (JSON output is non-interactive; progress is omitted).
- Diffs: Rendered as structured objects with
hunksarrays. - No ANSI codes in any structural element. Raw ANSI codes are preserved only in string values that represent verbatim terminal output.
- Pretty-printed by default (indented). Compact mode available via
<span style="color: cyan;">--json-compact</span>(future option). - Consistent schema per command: Each command's JSON schema is stable and documented, enabling reliable programmatic parsing.
Example (agents <span style="color: cyan;">--format</span> json project show local/api-service):
{
"command": "project show",
"status": "ok",
"exit_code": 0,
"data": {
"project": {
"name": "local/api-service",
"description": "Backend API",
"type": "local",
"remote": false,
"created_at": "2026-02-08T12:46:00Z"
},
"linked_resources": [
{
"name": "local/api-repo",
"type": "git-checkout",
"sandbox_strategy": "git_worktree",
"read_only": false
},
{
"name": "local/staging-db",
"type": "local/database",
"sandbox_strategy": "transaction_rollback",
"read_only": true
}
],
"validations": [
{
"name": "local/run-tests",
"command": "pytest --cov=src --cov-fail-under=80",
"mode": "required",
"timeout": 600,
"resource": "repo"
},
{
"name": "local/lint-check",
"command": "ruff check .",
"mode": "required",
"timeout": 300,
"resource": null
},
{
"name": "local/check-bundle-size",
"command": "node scripts/check-bundle-size.js",
"mode": "informational",
"timeout": 300,
"resource": null
}
],
"context": {
"include_resources": ["repo"],
"exclude_paths": ["**/node_modules/**"],
"max_file_size_bytes": 1048576
},
"indexing": {
"text_index": "ready",
"vector_index": "ready",
"graph_store": "disabled",
"indexed_files": 347,
"last_indexed_at": "2026-02-08T12:48:00Z"
},
"active_plans": [
{
"plan_id": "01HXM7A9",
"action": "local/code-coverage",
"phase": "execute"
}
]
},
"timing": {
"duration_ms": 42
},
"messages": [
{ "level": "ok", "text": "Project loaded" }
]
}
Example (agents --format json plan list --phase execute):
{
"command": "plan list",
"status": "ok",
"exit_code": 0,
"data": {
"plans": [
{
"id": "01HXM7A9",
"phase": "execute",
"state": "processing",
"action": "local/code-coverage",
"project": "local/api-service",
"elapsed": "00:01:12"
}
],
"filters": {
"phase": "execute",
"state": null,
"project": null,
"action": null
},
"summary": {
"total": 1,
"processing": 1,
"completed": 0,
"errored": 0
}
},
"timing": {
"duration_ms": 18
},
"messages": [
{ "level": "ok", "text": "1 plan listed" }
]
}
yaml — YAML Data Structure
Philosophy: Same data as json but in YAML format. Preferred by users who find YAML more readable for configuration and scripting workflows. Follows the same structural conventions as json.
Rendering rules:
- Same data envelope as JSON (
command,status,exit_code,data,timing,messages). - YAML 1.2 compliant output.
- Multi-line strings use YAML block scalars (
|for literal,>for folded) when appropriate. - No ANSI codes in structural elements (same rule as JSON).
- Sorted keys for deterministic output.
Example (agents --format yaml project show local/api-service):
command: project show
status: ok
exit_code: 0
data:
project:
name: local/api-service
description: Backend API
type: local
remote: false
created_at: "2026-02-08T12:46:00Z"
linked_resources:
- name: local/api-repo
type: git-checkout
sandbox_strategy: git_worktree
read_only: false
- name: local/staging-db
type: local/database
sandbox_strategy: transaction_rollback
read_only: true
validations:
- name: local/run-tests
command: "pytest --cov=src --cov-fail-under=80"
mode: required
timeout: 600
resource: repo
- name: local/lint-check
command: "ruff check ."
mode: required
timeout: 300
resource: null
- name: local/check-bundle-size
command: "node scripts/check-bundle-size.js"
mode: informational
timeout: 300
resource: null
context:
include_resources:
- repo
exclude_paths:
- "**/node_modules/**"
max_file_size_bytes: 1048576
indexing:
text_index: ready
vector_index: ready
graph_store: disabled
indexed_files: 347
last_indexed_at: "2026-02-08T12:48:00Z"
active_plans:
- plan_id: 01HXM7A9
action: local/code-coverage
phase: execute
timing:
duration_ms: 42
messages:
- level: ok
text: Project loaded
Format Comparison Matrix
| Capability | plain |
color |
table |
rich |
json |
yaml |
|---|---|---|---|---|---|---|
| Color codes | No | Yes | Yes | Yes | No | No |
| Box drawing | No | No | Yes | Yes | No | No |
| Animation/spinners | No | No | No | Yes | No | No |
| Live updates | No | No | No | Yes | No | No |
| Cursor movement | No | No | No | Yes | No | No |
| Syntax highlighting | No | No | No | Yes | No | No |
| Collapsible sections | No | No | No | Yes | No | No |
| Machine-parseable | Partially | Partially | No | No | Yes | Yes |
| Pipe-safe | Yes | No* | No | No | Yes | Yes |
| Unicode required | No | No | Yes | Yes | No | No |
| TTY required | No | No | No | Yes** | No | No |
* color output can be piped if the consumer understands ANSI codes (e.g., less -R).
** rich gracefully degrades to table when stdout is not a TTY.
Renderer Registration and Extension
The framework uses a registry pattern for format renderers, enabling third-party or plugin renderers. Each format is registered as a (MaterializationStrategy, ElementRenderer) pair — the strategy controls timing/ordering, and the renderer controls visual formatting:
@dataclass class FormatRegistration: """A registered format: its strategy factory, renderer factory, and fallback.""" strategy_factory: Callable[[TerminalCapabilities], MaterializationStrategy] renderer_factory: Callable[[TerminalCapabilities], ElementRenderer] fallback: str | None # Format name to fall back to, or Noneclass RendererRegistry: """Central registry for format (strategy, renderer) pairs. Formats are registered by name and resolved at runtime based on the active format and terminal capabilities. The registry supports dynamic registration, enabling plugins to add custom formats (e.g., 'html', 'csv', 'markdown'). Built-in registrations: "rich" → (LiveMaterializer, RichElementRenderer), fallback="table" "table" → (SequentialBufferMaterializer, TableElementRenderer), fallback="color" "color" → (SequentialBufferMaterializer, ColorElementRenderer), fallback="plain" "plain" → (SequentialBufferMaterializer, PlainElementRenderer), fallback=None "json" → (AccumulateMaterializer, JsonElementRenderer), fallback=None "yaml" → (AccumulateMaterializer, YamlElementRenderer), fallback=None """
_formats: <span style="color: cyan;">dict</span>[<span style="color: cyan;">str</span>, FormatRegistration] = {} <span style="color: yellow;">@classmethod</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">register</span>(cls, format_name: <span style="color: cyan;">str</span>, strategy_factory: <span style="color: cyan;">Callable</span>[[TerminalCapabilities], MaterializationStrategy], renderer_factory: <span style="color: cyan;">Callable</span>[[TerminalCapabilities], ElementRenderer], fallback: <span style="color: cyan;">str</span> | <span style="color: magenta; font-weight: 600;">None</span> = <span style="color: magenta; font-weight: 600;">None</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #66cc66;">"""Register a format.</span>Args: format_name: The format identifier (e.g., 'rich', 'json'). strategy_factory: Callable that creates a MaterializationStrategy, given terminal capabilities. renderer_factory: Callable that creates an ElementRenderer, given terminal capabilities. fallback: Optional fallback format name if this format cannot operate in the current terminal environment. """ cls._formats[format_name] = FormatRegistration( strategy_factory=strategy_factory, renderer_factory=renderer_factory, fallback=fallback, )
<span style="color: yellow;">@classmethod</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">resolve</span>(cls, format_name: <span style="color: cyan;">str</span>, terminal_caps: TerminalCapabilities ) -> <span style="color: cyan;">tuple</span>[MaterializationStrategy, ElementRenderer]: <span style="color: #66cc66;">"""Resolve the best (strategy, renderer) pair for the given format.</span>Walks the fallback chain if the requested format's renderer cannot operate in the current terminal environment. Returns the first pair where the renderer reports can_render(terminal_caps) == True. Raises: ValueError: If no usable format is found (should never happen since 'plain' has no fallback and always works). """ current = format_name visited: set[str] = set()
<span style="color: magenta; font-weight: 600;">while</span> current <span style="color: magenta; font-weight: 600;">and</span> current <span style="color: magenta; font-weight: 600;">not</span> <span style="color: magenta; font-weight: 600;">in</span> visited: visited.add(current) registration = cls._formats.get(current) <span style="color: magenta; font-weight: 600;">if</span> registration <span style="color: magenta; font-weight: 600;">is</span> <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: magenta; font-weight: 600;">break</span> renderer = registration.renderer_factory(terminal_caps) <span style="color: magenta; font-weight: 600;">if</span> renderer.can_render(terminal_caps): strategy = registration.strategy_factory(terminal_caps) strategy.bind(renderer, terminal_caps=terminal_caps) <span style="color: magenta; font-weight: 600;">return</span> strategy, renderer current = registration.fallback <span style="opacity: 0.7;"># Ultimate fallback is always plain</span> plain = cls._formats[<span style="color: #66cc66;">"plain"</span>] renderer = plain.renderer_factory(terminal_caps) strategy = plain.strategy_factory(terminal_caps) strategy.bind(renderer, terminal_caps=terminal_caps) <span style="color: magenta; font-weight: 600;">return</span> strategy, renderer <span style="color: yellow;">@classmethod</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">available_formats</span>(cls) -> <span style="color: cyan;">list</span>[<span style="color: cyan;">str</span>]: <span style="color: #66cc66;">"""Return all registered format names."""</span> <span style="color: magenta; font-weight: 600;">return</span> sorted(cls._formats.keys()) <span style="color: yellow;">@classmethod</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">is_registered</span>(cls, format_name: <span style="color: cyan;">str</span>) -> <span style="color: cyan;">bool</span>: <span style="color: #66cc66;">"""Check if a format is registered."""</span> <span style="color: magenta; font-weight: 600;">return</span> format_name <span style="color: magenta; font-weight: 600;">in</span> cls._formats
Terminal Capability Detection
The framework detects terminal capabilities to guide format resolution, strategy selection, and renderer fallback:
@dataclass class TerminalCapabilities: """Detected capabilities of the output terminal. This dataclass is populated once at CLI startup and passed to the RendererRegistry for format resolution. It is also available to individual strategies and renderers for fine-grained adaptation (e.g., adjusting column widths to terminal width, choosing between 256-color and truecolor palettes). """ is_tty: bool # Is stdout a TTY? width: int # Terminal width in columns height: int # Terminal height in rows supports_ansi: bool # Supports basic ANSI escape codes? supports_256_color: bool # Supports 256-color palette? supports_truecolor: bool # Supports 24-bit truecolor? supports_unicode: bool # Supports Unicode (box-drawing, etc.)? supports_cursor_movement: bool # Supports cursor repositioning? supports_alternate_screen: bool # Supports alternate screen buffer? no_color: bool # Is NO_COLOR environment variable set? term_program: str | None # TERM_PROGRAM value (e.g., "iTerm2", "vscode")<span style="color: yellow;">@classmethod</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">detect</span>(cls) -> <span style="color: #66cc66;">"TerminalCapabilities"</span>: <span style="color: #66cc66;">"""Auto-detect terminal capabilities from the environment.</span>
Detection logic: - is_tty: os.isatty(sys.stdout.fileno()) - width/height: os.get_terminal_size() with fallback to (80, 24) - supports_ansi: True if is_tty and not Windows legacy console - supports_256_color: True if TERM contains "256color" or COLORTERM is set - supports_truecolor: True if COLORTERM is "truecolor" or "24bit" - supports_unicode: True if locale encoding is UTF-8 - supports_cursor_movement: True if is_tty and TERM is not "dumb" - supports_alternate_screen: True if supports_cursor_movement - no_color: True if NO_COLOR environment variable is set (any value) - term_program: Value of TERM_PROGRAM environment variable """ ...
Plugin Format Registration
Third-party plugins can register custom formats using the registry:
# Example: Registering a custom 'csv' format plugin from cleveragents.output import RendererRegistry, SequentialBufferMaterializerclass CsvElementRenderer(ElementRenderer): """Renders tables as CSV, other elements as plain text.""" format_name = "csv"
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">render_table</span>(self, table: Table, stream: <span style="color: cyan;">IO</span>) -> <span style="color: magenta; font-weight: 600;">None</span>: writer = csv.writer(stream) writer.writerow([col.name <span style="color: magenta; font-weight: 600;">for</span> col <span style="color: magenta; font-weight: 600;">in</span> table.columns]) <span style="color: magenta; font-weight: 600;">for</span> row <span style="color: magenta; font-weight: 600;">in</span> table.rows: writer.writerow([row.get(col.name, <span style="color: #66cc66;">""</span>) <span style="color: magenta; font-weight: 600;">for</span> col <span style="color: magenta; font-weight: 600;">in</span> table.columns]) <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">can_render</span>(self, terminal_caps: TerminalCapabilities) -> <span style="color: cyan;">bool</span>: <span style="color: magenta; font-weight: 600;">return</span> <span style="color: magenta; font-weight: 600;">True</span> <span style="opacity: 0.7;"># CSV works everywhere</span>
# Register at plugin load time RendererRegistry.register( format_name="csv", strategy_factory=lambda caps: SequentialBufferMaterializer(), renderer_factory=lambda caps: CsvElementRenderer(), fallback="plain", )
Edge Cases and Special Behaviors
Error Output
Errors are rendered through the same framework. When a command raises an exception or signals an error, the session's context manager (__exit__) catches the exception, creates a StatusMessage with level="error" and an optional TextBlock with details, and closes the session with exit_code=1. In json/yaml formats, errors produce:
{
"command": "project show",
"status": "error",
"exit_code": 1,
"error": {
"code": "NOT_FOUND",
"message": "Project 'local/nonexistent' not found",
"detail": "No project with name 'local/nonexistent' exists. Run 'agents project list' to see available projects.",
"suggestions": [
"agents project list",
"agents project create local/nonexistent"
]
},
"timing": { "duration_ms": 5 }
}
Empty Results
When a list command returns no results, all formats handle it gracefully:
- plain/color: Prints a message like
No projects found. - table: Renders an empty table with headers and a
(empty)message. - rich: Renders a dimmed panel with an empty-state message and suggested actions.
- json/yaml: Returns an empty array in the appropriate data field.
Large Data Sets
For commands that may return large result sets (e.g., resource list with thousands of resources):
- plain/color/table (SequentialBufferMaterializer): The table handle accumulates rows as the producer adds them. Since the buffer is in-memory, very large result sets benefit from the
max_rows_hint— the renderer truncates display at N rows with a(... and N more rows)indicator. The full data is still available in the snapshot for json/yaml. - rich (LiveMaterializer): Uses a virtual-scrolling table that renders only visible rows. Shows a count indicator (e.g., "Showing 1-50 of 2,847"). New rows animate into view as they are added.
- json/yaml (AccumulateMaterializer): Accumulates and emits a complete array at session end. Streaming JSON lines (one JSON object per row) for very large sets is a future consideration.
Nested/Recursive Structures
Tree-like data (resource trees, decision trees, plan hierarchies) may be arbitrarily deep. Renderers respect max_depth_hint:
- plain/color: Truncate at depth N with a
... (N more levels)indicator. - table: Same truncation, rendered inside a box.
- rich: Collapsible tree nodes — deep levels start collapsed. User can expand interactively if the terminal supports it.
- json/yaml: Full depth — no truncation (programmatic consumers need complete data).
Mixed Content
Some commands produce mixed output (e.g., plan status has panels, tables, progress bars, and status messages). Each element is created via its own handle on the session, and the materialization strategy renders them in declaration order. The ElementRenderer is responsible for visual spacing and grouping between heterogeneous elements (e.g., inserting blank lines between a panel and a table in plain format, or adding visual margins in rich format).
Producer Error Mid-Stream
When a producer encounters an error while writing to a handle (e.g., an API call fails while populating a table), the framework handles it as follows:
-
The handle is closed with partial data — the producer catches its exception, optionally calls
handle.close()(or lets the context manager close it), and then creates aStatusMessagehandle withlevel="error"to report the failure. -
The materialization strategy renders whatever was accumulated — a table with 3 of an expected 10 rows is rendered with those 3 rows, followed by the error message. This is better than rendering nothing.
-
The session's exit code is set to 1 — indicating partial failure.
-
For
json/yamlformats, the accumulated snapshot includes both the partial data and the error message in themessagesarray, giving programmatic consumers full visibility.
Example of producer error handling:
async def list_resources(session: OutputSession, client: ApiClient) -> None: table = session.table("Resources", columns=[ ColumnDef(name="Name"), ColumnDef(name="Type"), ColumnDef(name="Status"), ])<span style="color: magenta; font-weight: 600;">try</span>: <span style="color: magenta; font-weight: 600;">async</span> <span style="color: magenta; font-weight: 600;">for</span> resource <span style="color: magenta; font-weight: 600;">in</span> client.list_resources(): table.add_row({ <span style="color: #66cc66;">"Name"</span>: resource.name, <span style="color: #66cc66;">"Type"</span>: resource.type, <span style="color: #66cc66;">"Status"</span>: resource.status, }) <span style="color: magenta; font-weight: 600;">except</span> ApiError <span style="color: magenta; font-weight: 600;">as</span> e: table.close() <span style="opacity: 0.7;"># Close with partial data</span> session.status(<span style="color: #66cc66;">f"Error fetching resources: {e}"</span>, level=<span style="color: #66cc66;">"error"</span>) <span style="color: magenta; font-weight: 600;">return</span> table.close() session.status(<span style="color: #66cc66;">f"{table.element.row_count} resources listed"</span>, level=<span style="color: #66cc66;">"ok"</span>)
Abandoned Handles
If a handle is never explicitly closed and the session ends (either normally via session.close() or via the context manager's __exit__), the session force-closes all remaining open handles:
- A warning is logged (not rendered to the user):
"Handle {handle_id} ({element_type}) was not explicitly closed; force-closing at session end." - The handle is closed with its current accumulated state.
- The materialization strategy processes the
ElementClosedevent normally.
This ensures that no data is silently lost, even if producer code forgets to close a handle due to an unhandled code path.
Back-Pressure and Throttling
The LiveMaterializer (used by rich format) limits terminal redraws to its configured frame rate (default 15 fps). When handles emit updates faster than the frame rate:
- Updates are coalesced — the materializer tracks a dirty set of handle IDs that have been updated since the last frame.
- On each frame tick, all dirty elements are redrawn in a single pass, and the dirty set is cleared.
- The event queue between session and strategy uses bounded capacity. If the queue fills (producer is vastly faster than rendering), the session's
_emit_eventmethod dropsElementUpdatedevents for handles that are already in the dirty set (since the next frame will redraw them anyway).ElementCreatedandElementClosedevents are never dropped.
For SequentialBufferMaterializer and AccumulateMaterializer, there is no back-pressure concern — updates are buffered in memory and never rendered incrementally.
Cancellation Semantics
When a command is cancelled (e.g., user presses Ctrl+C):
- The session receives a cancellation signal and enters the
"closing"state. - All open handles are force-closed with their current state.
- A
StatusMessagewithlevel="warn"and message"Operation cancelled"is emitted. - The session closes with
exit_code=130(standard SIGINT exit code). - For
richformat: TheLiveMaterializerfreezes the display, resolves any active spinners to a cancellation indicator (e.g., yellow⚠), and moves the cursor to the end of the output. - For buffered formats: Any elements that have been rendered stay on screen. Pending buffered elements are flushed in order, followed by the cancellation message.
- For
json/yaml: The accumulated snapshot is serialized with"status": "cancelled"and the appropriate exit code.
Interleaved Status Messages During Concurrent Production
When multiple producers are running concurrently (e.g., populating two tables), status messages may be created at any time by any producer. The materialization strategy handles these based on format:
rich(LiveMaterializer): Status messages are rendered immediately in a dedicated status region at the bottom of the display (below all element regions). Multiple concurrent status messages stack vertically.plain/color/table(SequentialBufferMaterializer): Status messages created during concurrent production are treated as elements in declaration order, just like tables and panels. A status message created between two table creations will be rendered between those tables. Status messages created after all tables will render after all tables are flushed.json/yaml(AccumulateMaterializer): All status messages are collected in themessagesarray of the final snapshot, ordered by timestamp.
Integration with Future TUI
The reactive OutputSession architecture is intentionally designed to serve as the data layer for a future TUI (text user interface). The session's event-driven model maps directly to TUI widget patterns:
-
Element handles become observable data sources. A TUI
MaterializationStrategy(e.g.,TuiMaterializer) would subscribe to element events and route them to TUI widgets. The producer code (command logic) is completely unaware of whether it is driving a CLI, TUI, or web frontend — it writes to handles identically in all cases. -
Element types map to TUI widgets:
PanelHandle→ info pane or detail card widgetTableHandle→ sortable, filterable data grid widget (rows arrive incrementally viaadd_rowevents)TreeHandle→ collapsible tree view widget (nodes arrive incrementally viaadd_childevents)ProgressHandle→ animated progress bar or step-list widgetStatusHandle→ toast notification or status bar messageCodeHandle→ syntax-highlighted code viewer widgetDiffHandle→ side-by-side diff viewer widget
-
Interactive features are additive. The TUI can offer features that the CLI cannot — sorting table columns, filtering rows, collapsing/expanding tree nodes, searching within code blocks — without any changes to producer code. These features are implemented in the TUI's
ElementRendererand widget layer. -
Concurrent updates are native. Because the session already supports multiple concurrent producers writing to different handles, the TUI naturally displays multiple simultaneously-updating widgets (e.g., two tables being populated in parallel by concurrent operations). The
TuiMaterializerroutes events to widgets, and each widget redraws independently using the TUI framework's event loop. -
The
StructuredOutputsnapshot provides the initial state when navigating to a completed session in the TUI (e.g., reviewing a past command's output), while live sessions use the event stream.
The separation between production (element handles), timing (materialization strategy), and presentation (element renderer) ensures that the same command logic supports CLI, TUI, and web frontends without modification — only the (MaterializationStrategy, ElementRenderer) pair changes.
Programmatic Usage Examples
This section demonstrates how command implementations use the Output Rendering Framework through the OutputSession API, and how the same producer code produces correct output across all formats.
Example 1: Simple Static Command Output
The simplest usage — a command that creates elements, populates them synchronously, and closes them. No concurrency, no streaming.
Producer code (agents project show):
async def cmd_project_show(session: OutputSession, project_name: str) -> None: """Implementation of 'agents project show <project>'.""" project = await api.get_project(project_name) resources = await api.list_project_resources(project.name) validations = await api.list_project_validations(project.name)<span style="opacity: 0.7;"># --- Build output elements ---</span> <span style="opacity: 0.7;"># Panel: Project details</span> <span style="color: magenta; font-weight: 600;">with</span> session.panel(<span style="color: #66cc66;">"Project Details"</span>) <span style="color: magenta; font-weight: 600;">as</span> panel: panel.set_entries({ <span style="color: #66cc66;">"Name"</span>: project.name, <span style="color: #66cc66;">"Description"</span>: project.description, <span style="color: #66cc66;">"Resources"</span>: <span style="color: cyan;">str</span>(len(resources)), <span style="color: #66cc66;">"Remote"</span>: <span style="color: #66cc66;">"yes"</span> <span style="color: magenta; font-weight: 600;">if</span> project.remote <span style="color: magenta; font-weight: 600;">else</span> <span style="color: #66cc66;">"no"</span>, <span style="color: #66cc66;">"Created"</span>: project.created_at.strftime(<span style="color: #66cc66;">"%Y-%m-%d %H:%M"</span>), }, style_hints={ <span style="color: #66cc66;">"Name"</span>: <span style="color: #66cc66;">"identifier"</span>, <span style="color: #66cc66;">"Resources"</span>: <span style="color: #66cc66;">"number"</span>, <span style="color: #66cc66;">"Remote"</span>: <span style="color: #66cc66;">"success"</span> <span style="color: magenta; font-weight: 600;">if</span> <span style="color: magenta; font-weight: 600;">not</span> project.remote <span style="color: magenta; font-weight: 600;">else</span> <span style="color: #66cc66;">"info"</span>, <span style="color: #66cc66;">"Created"</span>: <span style="color: #66cc66;">"success"</span>, }) <span style="opacity: 0.7;"># Table: Linked resources</span> <span style="color: magenta; font-weight: 600;">with</span> session.table(<span style="color: #66cc66;">"Linked Resources"</span>, columns=[ ColumnDef(name=<span style="color: #66cc66;">"Resource"</span>, type=<span style="color: #66cc66;">"string"</span>, style_hint=<span style="color: #66cc66;">"identifier"</span>), ColumnDef(name=<span style="color: #66cc66;">"Type"</span>), ColumnDef(name=<span style="color: #66cc66;">"Sandbox"</span>), ColumnDef(name=<span style="color: #66cc66;">"Read-Only"</span>), ]) <span style="color: magenta; font-weight: 600;">as</span> table: <span style="color: magenta; font-weight: 600;">for</span> r <span style="color: magenta; font-weight: 600;">in</span> resources: table.add_row({ <span style="color: #66cc66;">"Resource"</span>: r.name, <span style="color: #66cc66;">"Type"</span>: r.type, <span style="color: #66cc66;">"Sandbox"</span>: r.sandbox_strategy, <span style="color: #66cc66;">"Read-Only"</span>: <span style="color: #66cc66;">"yes"</span> <span style="color: magenta; font-weight: 600;">if</span> r.read_only <span style="color: magenta; font-weight: 600;">else</span> <span style="color: #66cc66;">"no"</span>, }) <span style="opacity: 0.7;"># Table: Validations</span> <span style="color: magenta; font-weight: 600;">with</span> session.table(<span style="color: #66cc66;">f"Validations ({len(validations)})"</span>, columns=[ ColumnDef(name=<span style="color: #66cc66;">"ID"</span>, type=<span style="color: #66cc66;">"id"</span>, style_hint=<span style="color: #66cc66;">"identifier"</span>), ColumnDef(name=<span style="color: #66cc66;">"Command"</span>), ColumnDef(name=<span style="color: #66cc66;">"Mode"</span>), ]) <span style="color: magenta; font-weight: 600;">as</span> table: <span style="color: magenta; font-weight: 600;">for</span> v <span style="color: magenta; font-weight: 600;">in</span> validations: table.add_row({ <span style="color: #66cc66;">"ID"</span>: v.id, <span style="color: #66cc66;">"Command"</span>: v.command, <span style="color: #66cc66;">"Mode"</span>: v.mode, }) <span style="opacity: 0.7;"># Status: Final message</span> session.status(<span style="color: #66cc66;">"Project loaded"</span>, level=<span style="color: #66cc66;">"ok"</span>)
What this produces in plain format:
Project Details Name: local/api-service Description: Backend API Resources: 2 Remote: no Created: 2026-02-08 12:46Linked Resources Resource Type Sandbox Read-Only
local/api-repo git-checkout git_worktree no local/staging-db local/database transaction_rollback yes
Validations (3) Name Command Mode
local/run-tests pytest --cov=src --cov-fail-under=80 required local/lint-check ruff check . required local/check-bundle-size node scripts/check-bundle-size.js informational
[OK] Project loaded
What this produces in rich format:
╭─ Project Details ──────────────╮ │ Name: local/api-service │ │ Description: Backend API │ │ Resources: 2 │ │ Remote: no │ │ Created: 2026-02-08 12:46 │ ╰────────────────────────────────╯╭─ Linked Resources ──────────────────────────────────────────────────────╮ │ Resource Type Sandbox Read-Only │ │ ──────────────── ────────────── ──────────────────── ───────── │ │ local/api-repo git-checkout git_worktree no │ │ local/staging-db local/database transaction_rollback yes │ ╰─────────────────────────────────────────────────────────────────────────╯
╭─ Validations (3) ──────────────────────────────────────────────────────────────╮ │ Name Command Mode │ │ ─────────────────────── ───────────────────────────────────── ─────────── │ │ local/run-tests pytest --cov=src --cov-fail-under=80 required │ │ local/lint-check ruff check . required │ │ local/check-bundle-size node scripts/check-bundle-size.js informational │ ╰────────────────────────────────────────────────────────────────────────────────╯
✓ OK Project loaded
What this produces in json format:
{
"command": "project show",
"status": "ok",
"exit_code": 0,
"data": {
"project_details": {
"Name": "local/api-service",
"Description": "Backend API",
"Resources": "2",
"Remote": "no",
"Created": "2026-02-08 12:46"
},
"linked_resources": [
{
"Resource": "local/api-repo",
"Type": "git-checkout",
"Sandbox": "git_worktree",
"Read-Only": "no"
},
{
"Resource": "local/staging-db",
"Type": "local/database",
"Sandbox": "transaction_rollback",
"Read-Only": "yes"
}
],
"validations": [
{
"Name": "local/run-tests",
"Command": "pytest --cov=src --cov-fail-under=80",
"Mode": "required"
},
{
"Name": "local/lint-check",
"Command": "ruff check .",
"Mode": "required"
},
{
"Name": "local/check-bundle-size",
"Command": "node scripts/check-bundle-size.js",
"Mode": "informational"
}
]
},
"timing": { "duration_ms": 42 },
"messages": [
{ "level": "ok", "text": "Project loaded" }
]
}
In all three formats, the producer code is identical. The OutputSession and its materialization strategy handle the differences transparently.
Example 2: Streaming Rows into a Table
A command that streams rows into a table as results arrive from a paginated API. The table handle stays open while the producer fetches pages.
Producer code (agents resource list):
async def cmd_resource_list(session: OutputSession, project: str | None) -> None: """Implementation of 'agents resource list'."""<span style="opacity: 0.7;"># Create the table handle — it will accumulate rows as we stream them</span> table = session.table(<span style="color: #66cc66;">"Resources"</span>, columns=[ ColumnDef(name=<span style="color: #66cc66;">"Name"</span>, type=<span style="color: #66cc66;">"string"</span>, style_hint=<span style="color: #66cc66;">"identifier"</span>), ColumnDef(name=<span style="color: #66cc66;">"Type"</span>), ColumnDef(name=<span style="color: #66cc66;">"Project"</span>), ColumnDef(name=<span style="color: #66cc66;">"Sandbox"</span>), ColumnDef(name=<span style="color: #66cc66;">"Status"</span>), ], sort_key=<span style="color: #66cc66;">"Name"</span>) <span style="opacity: 0.7;"># Create a progress indicator for the fetch operation</span> progress = session.progress(<span style="color: #66cc66;">"Fetching resources..."</span>, indeterminate=<span style="color: magenta; font-weight: 600;">True</span>) <span style="opacity: 0.7;"># Stream pages from the API</span> count = 0 <span style="color: magenta; font-weight: 600;">async</span> <span style="color: magenta; font-weight: 600;">for</span> page <span style="color: magenta; font-weight: 600;">in</span> api.list_resources_paginated(project=project): <span style="color: magenta; font-weight: 600;">for</span> resource <span style="color: magenta; font-weight: 600;">in</span> page.items: table.add_row({ <span style="color: #66cc66;">"Name"</span>: resource.name, <span style="color: #66cc66;">"Type"</span>: resource.type, <span style="color: #66cc66;">"Project"</span>: resource.project, <span style="color: #66cc66;">"Sandbox"</span>: resource.sandbox_strategy, <span style="color: #66cc66;">"Status"</span>: resource.status, }) count += 1 <span style="opacity: 0.7;"># Update progress label with count so far</span> progress.set_label(<span style="color: #66cc66;">f"Fetching resources... ({count} found)"</span>) <span style="opacity: 0.7;"># Close the progress indicator (it has served its purpose)</span> progress.close() <span style="opacity: 0.7;"># Set summary and close the table</span> table.set_summary({<span style="color: #66cc66;">"total"</span>: count}) table.close() <span style="opacity: 0.7;"># Final status</span> session.status(<span style="color: #66cc66;">f"{count} resources listed"</span>, level=<span style="color: #66cc66;">"ok"</span>)
What this looks like in plain format (SequentialBufferMaterializer):
The progress indicator is rendered as a static line. The table is buffered until table.close() is called, then rendered in full. The user sees nothing until the fetch is complete — then the entire result appears at once:
Fetching resources... (47 found) [done]Resources Name Type Project Sandbox Status
local/api-repo git-checkout local/api-service git_worktree active local/staging-db local/database local/api-service transaction_rollback active local/docs-repo git-checkout local/docs-site git_worktree active ... (44 more rows)
Total: 47
[OK] 47 resources listed
What this looks like in rich format (LiveMaterializer):
The progress spinner animates in real-time. The table updates in-place as rows arrive — each new row appears at the bottom of the table, the row count updates, and the terminal display is rewritten without scrolling. This is a static snapshot of the live display mid-fetch:
⠙ Fetching resources... (23 found)
╭─ Resources ────────────────────────────────────────────────────────────────────────────╮ │ Name Type Project Sandbox Status │ │ ──────────────────── ────────────── ───────────────── ──────────────────── ────── │ │ local/api-repo git-checkout local/api-service git_worktree active │ │ local/staging-db local/database local/api-service transaction_rollback active │ │ local/docs-repo git-checkout local/docs-site git_worktree active │ │ ... │ │ local/test-fixtures git-checkout local/api-service git_worktree active │ │ │ │ Showing 1-23 of 23 (fetching...) │ ╰────────────────────────────────────────────────────────────────────────────────────────╯
In rich mode, the spinner animates, the table grows as rows arrive, and the row count updates — all in-place without scrolling. When the fetch completes, the spinner resolves to ✓, and the table shows its final state.
Example 3: Concurrent Parallel Operations (Two Tables Simultaneously)
This is the key motivating example for the reactive architecture. Two tables are populated simultaneously by parallel workers, and the producer code is completely format-agnostic.
Producer code (agents plan status — showing resources and active tool calls concurrently):
async def cmd_plan_status(session: OutputSession, plan_id: str) -> None: """Implementation of 'agents plan status <plan_id>'. This command fetches plan metadata, then concurrently streams two data sources: resource statuses and active tool call logs. Both data sources are long-running — they produce results over several seconds as the backend resolves each item. """ plan = await api.get_plan(plan_id)<span style="opacity: 0.7;"># Panel: Plan metadata (created and closed synchronously)</span> <span style="color: magenta; font-weight: 600;">with</span> session.panel(<span style="color: #66cc66;">"Plan"</span>) <span style="color: magenta; font-weight: 600;">as</span> panel: panel.set_entries({ <span style="color: #66cc66;">"Plan ID"</span>: plan.id, <span style="color: #66cc66;">"Phase"</span>: plan.phase, <span style="color: #66cc66;">"State"</span>: plan.state, <span style="color: #66cc66;">"Action"</span>: plan.action, <span style="color: #66cc66;">"Project"</span>: plan.project, <span style="color: #66cc66;">"Started"</span>: plan.started_at.strftime(<span style="color: #66cc66;">"%H:%M:%S"</span>), }, style_hints={ <span style="color: #66cc66;">"Plan ID"</span>: <span style="color: #66cc66;">"identifier"</span>, <span style="color: #66cc66;">"Phase"</span>: <span style="color: #66cc66;">"info"</span>, <span style="color: #66cc66;">"State"</span>: <span style="color: #66cc66;">"warning"</span> <span style="color: magenta; font-weight: 600;">if</span> plan.state == <span style="color: #66cc66;">"processing"</span> <span style="color: magenta; font-weight: 600;">else</span> <span style="color: #66cc66;">"success"</span>, }) <span style="opacity: 0.7;"># Create both table handles BEFORE starting concurrent producers.</span> <span style="opacity: 0.7;"># Declaration order determines rendering order in sequential formats.</span> resource_table = session.table(<span style="color: #66cc66;">"Resource Status"</span>, columns=[ ColumnDef(name=<span style="color: #66cc66;">"Resource"</span>, style_hint=<span style="color: #66cc66;">"identifier"</span>), ColumnDef(name=<span style="color: #66cc66;">"Type"</span>), ColumnDef(name=<span style="color: #66cc66;">"Status"</span>), ColumnDef(name=<span style="color: #66cc66;">"Latency"</span>, type=<span style="color: #66cc66;">"string"</span>, alignment=<span style="color: #66cc66;">"right"</span>), ]) tool_table = session.table(<span style="color: #66cc66;">"Tool Call Log"</span>, columns=[ ColumnDef(name=<span style="color: #66cc66;">"#"</span>, type=<span style="color: #66cc66;">"number"</span>, alignment=<span style="color: #66cc66;">"right"</span>), ColumnDef(name=<span style="color: #66cc66;">"Tool"</span>), ColumnDef(name=<span style="color: #66cc66;">"Target"</span>), ColumnDef(name=<span style="color: #66cc66;">"Result"</span>), ColumnDef(name=<span style="color: #66cc66;">"Duration"</span>, type=<span style="color: #66cc66;">"string"</span>, alignment=<span style="color: #66cc66;">"right"</span>), ]) <span style="opacity: 0.7;"># --- Run two producers concurrently ---</span> <span style="opacity: 0.7;"># Each producer writes to its own handle. Neither producer knows</span> <span style="opacity: 0.7;"># which format is active. The materialization strategy handles</span> <span style="opacity: 0.7;"># the coordination.</span> <span style="color: magenta; font-weight: 600;">async</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">stream_resources</span>(): <span style="color: #66cc66;">"""Producer A: streams resource status checks."""</span> <span style="color: magenta; font-weight: 600;">async</span> <span style="color: magenta; font-weight: 600;">for</span> status <span style="color: magenta; font-weight: 600;">in</span> api.stream_resource_statuses(plan.id): resource_table.add_row({ <span style="color: #66cc66;">"Resource"</span>: status.resource_name, <span style="color: #66cc66;">"Type"</span>: status.resource_type, <span style="color: #66cc66;">"Status"</span>: status.status, <span style="color: #66cc66;">"Latency"</span>: <span style="color: #66cc66;">f"{status.latency_ms}ms"</span>, }) resource_table.close() <span style="color: magenta; font-weight: 600;">async</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">stream_tool_calls</span>(): <span style="color: #66cc66;">"""Producer B: streams tool call results."""</span> <span style="color: magenta; font-weight: 600;">async</span> <span style="color: magenta; font-weight: 600;">for</span> call <span style="color: magenta; font-weight: 600;">in</span> api.stream_tool_calls(plan.id): tool_table.add_row({ <span style="color: #66cc66;">"#"</span>: call.sequence_number, <span style="color: #66cc66;">"Tool"</span>: call.tool_name, <span style="color: #66cc66;">"Target"</span>: call.target, <span style="color: #66cc66;">"Result"</span>: call.result_summary, <span style="color: #66cc66;">"Duration"</span>: <span style="color: #66cc66;">f"{call.duration_ms}ms"</span>, }) tool_table.close() <span style="opacity: 0.7;"># Launch both producers concurrently</span> <span style="color: magenta; font-weight: 600;">await</span> asyncio.gather(stream_resources(), stream_tool_calls()) <span style="opacity: 0.7;"># Final status</span> session.status(<span style="color: #66cc66;">f"Plan {plan_id} status retrieved"</span>, level=<span style="color: #66cc66;">"ok"</span>)
What this produces in plain format (SequentialBufferMaterializer):
Both tables are populated concurrently, but the materializer buffers each one and renders them in declaration order when their handles close. The user sees nothing until the first-declared table (Resource Status) closes, then it prints. Then when the second table (Tool Call Log) closes, it prints. Data may have arrived interleaved across both tables, but the output is perfectly sequential:
Plan Plan ID: 01HXM7A9 Phase: execute State: processing Action: local/code-coverage Project: local/api-service Started: 12:58:10Resource Status Resource Type Status Latency
local/api-repo git-checkout ready 42ms local/staging-db local/database ready 128ms
Tool Call Log
Tool Target Result Duration
1 read_file src/auth/__init__.py 200 lines 0.1s 2 read_file src/auth/session.py 340 lines 0.1s 3 write_file tests/test_auth.py created 0.2s 4 edit_file src/auth/session.py 12 lines +/- 0.3s 5 run_tests pytest tests/test_auth 3 passed 2.1s
[OK] Plan 01HXM7A9 status retrieved
What this produces in rich format (LiveMaterializer):
Both tables are visible simultaneously and update in-place as data arrives. This snapshot shows the display mid-stream — the resource table has two rows and the tool call table has three so far:
╭─ Plan ──────────────────────────────╮ │ Plan ID: 01HXM7A9 │ │ Phase: execute │ │ State: processing │ │ Action: local/code-coverage │ │ Project: local/api-service │ │ Started: 12:58:10 │ ╰─────────────────────────────────────╯╭─ Resource Status ⠙ ───────────────────────────────────────────╮ │ Resource Type Status Latency │ │ ──────────────── ────────────── ─────── ─────── │ │ local/api-repo git-checkout ready 42ms │ │ local/staging-db local/database ready 128ms │ │ │ │ 2 resources (streaming...) │ ╰───────────────────────────────────────────────────────────────╯
╭─ Tool Call Log ⠙ ────────────────────────────────────────────────────╮ │ # Tool Target Result Duration │ │ ── ────────── ────────────────────── ───────────── ──────── │ │ 1 read_file src/auth/__init__.py 200 lines 0.1s │ │ 2 read_file src/auth/session.py 340 lines 0.1s │ │ 3 write_file tests/test_auth.py created 0.2s │ │ │ │ 3 calls (streaming...) │ ╰──────────────────────────────────────────────────────────────────────╯
In rich mode, both tables have animated spinners in their titles indicating active streaming. As new rows arrive from either producer, the corresponding table's display is updated in-place. When a producer finishes and closes its handle, the spinner resolves to a ✓ and the "(streaming...)" indicator is removed. The other table continues updating independently.
What this produces in json format (AccumulateMaterializer):
Nothing is printed until the session closes. Then the complete accumulated state is serialized:
{
"command": "plan status",
"status": "ok",
"exit_code": 0,
"data": {
"plan": {
"Plan ID": "01HXM7A9",
"Phase": "execute",
"State": "processing",
"Action": "local/code-coverage",
"Project": "local/api-service",
"Started": "12:58:10"
},
"resource_status": [
{ "Resource": "local/api-repo", "Type": "git-checkout", "Status": "ready", "Latency": "42ms" },
{ "Resource": "local/staging-db", "Type": "local/database", "Status": "ready", "Latency": "128ms" }
],
"tool_call_log": [
{ "#": 1, "Tool": "read_file", "Target": "src/auth/__init__.py", "Result": "200 lines", "Duration": "0.1s" },
{ "#": 2, "Tool": "read_file", "Target": "src/auth/session.py", "Result": "340 lines", "Duration": "0.1s" },
{ "#": 3, "Tool": "write_file", "Target": "tests/test_auth.py", "Result": "created", "Duration": "0.2s" },
{ "#": 4, "Tool": "edit_file", "Target": "src/auth/session.py", "Result": "12 lines +/-", "Duration": "0.3s" },
{ "#": 5, "Tool": "run_tests", "Target": "pytest tests/test_auth", "Result": "3 passed", "Duration": "2.1s" }
]
},
"timing": { "duration_ms": 3200 },
"messages": [
{ "level": "ok", "text": "Plan 01HXM7A9 status retrieved" }
]
}
The critical point: the producer code in all three formats is exactly the same. The asyncio.gather call runs both producers concurrently regardless of format. The materialization strategy — LiveMaterializer, SequentialBufferMaterializer, or AccumulateMaterializer — transparently decides how that concurrent data reaches the user.
Example 4: Progress with Concurrent Sub-Operations
A command that executes a multi-step process with a progress indicator, where some steps involve parallel sub-operations.
Producer code (agents plan execute):
async def cmd_plan_execute(session: OutputSession, plan_id: str) -> None: """Implementation of 'agents plan execute <plan_id>'.""" plan = await api.get_plan(plan_id)<span style="opacity: 0.7;"># Panel: Execution metadata</span> <span style="color: magenta; font-weight: 600;">with</span> session.panel(<span style="color: #66cc66;">"Execution"</span>) <span style="color: magenta; font-weight: 600;">as</span> panel: panel.set_entries({ <span style="color: #66cc66;">"Plan"</span>: plan.id, <span style="color: #66cc66;">"Phase"</span>: <span style="color: #66cc66;">"execute"</span>, <span style="color: #66cc66;">"Sandbox"</span>: plan.sandbox_strategy, <span style="color: #66cc66;">"Worker"</span>: plan.worker, <span style="color: #66cc66;">"Started"</span>: <span style="color: cyan;">datetime</span>.now().strftime(<span style="color: #66cc66;">"%H:%M:%S"</span>), }) <span style="opacity: 0.7;"># Progress indicator with named steps</span> progress = session.progress(<span style="color: #66cc66;">"Executing plan"</span>, total=4, steps=[ <span style="color: #66cc66;">"Collect context"</span>, <span style="color: #66cc66;">"Run tools"</span>, <span style="color: #66cc66;">"Build changeset"</span>, <span style="color: #66cc66;">"Validate"</span>, ]) <span style="opacity: 0.7;"># Step 1: Collect context</span> progress.set_step_status(<span style="color: #66cc66;">"Collect context"</span>, <span style="color: #66cc66;">"active"</span>) context = <span style="color: magenta; font-weight: 600;">await</span> api.collect_context(plan.id) progress.set_step_status(<span style="color: #66cc66;">"Collect context"</span>, <span style="color: #66cc66;">"done"</span>) progress.set_progress(1, 4) <span style="opacity: 0.7;"># Step 2: Run tools (parallel sub-operations)</span> progress.set_step_status(<span style="color: #66cc66;">"Run tools"</span>, <span style="color: #66cc66;">"active"</span>) tool_results = <span style="color: magenta; font-weight: 600;">await</span> api.run_tools(plan.id, context) progress.set_step_status(<span style="color: #66cc66;">"Run tools"</span>, <span style="color: #66cc66;">"done"</span>) progress.set_progress(2, 4) <span style="opacity: 0.7;"># Step 3: Build changeset</span> progress.set_step_status(<span style="color: #66cc66;">"Build changeset"</span>, <span style="color: #66cc66;">"active"</span>) changeset = <span style="color: magenta; font-weight: 600;">await</span> api.build_changeset(plan.id, tool_results) progress.set_step_status(<span style="color: #66cc66;">"Build changeset"</span>, <span style="color: #66cc66;">"done"</span>) progress.set_progress(3, 4) <span style="opacity: 0.7;"># Step 4: Validate</span> progress.set_step_status(<span style="color: #66cc66;">"Validate"</span>, <span style="color: #66cc66;">"active"</span>) validation = <span style="color: magenta; font-weight: 600;">await</span> api.validate_changeset(plan.id, changeset) progress.set_step_status(<span style="color: #66cc66;">"Validate"</span>, <span style="color: #66cc66;">"done"</span>) progress.set_progress(4, 4) progress.close() <span style="opacity: 0.7;"># Summary panel</span> <span style="color: magenta; font-weight: 600;">with</span> session.panel(<span style="color: #66cc66;">"Strategy Summary"</span>) <span style="color: magenta; font-weight: 600;">as</span> panel: panel.set_entries({ <span style="color: #66cc66;">"Decisions"</span>: <span style="color: cyan;">str</span>(changeset.decision_count), <span style="color: #66cc66;">"Invariants"</span>: <span style="color: cyan;">str</span>(changeset.invariant_count), <span style="color: #66cc66;">"Planned Child Plans"</span>: <span style="color: #66cc66;">f"{changeset.child_plan_count}+"</span>, <span style="color: #66cc66;">"Estimated Files"</span>: <span style="color: #66cc66;">f"~{changeset.file_count}"</span>, <span style="color: #66cc66;">"Risk"</span>: changeset.risk_level, }) <span style="opacity: 0.7;"># Final status</span> <span style="color: magenta; font-weight: 600;">if</span> validation.passed: session.status(<span style="color: #66cc66;">"Execution complete — all validations passed"</span>, level=<span style="color: #66cc66;">"ok"</span>) <span style="color: magenta; font-weight: 600;">else</span>: session.status( <span style="color: #66cc66;">f"Execution complete — {validation.failure_count} validation(s) failed"</span>, level=<span style="color: #66cc66;">"warn"</span>, detail=validation.summary, )
What this looks like in plain format:
The progress indicator renders as a static step list. Since SequentialBufferMaterializer buffers each element until its handle closes, the progress indicator is not visible during execution — it appears as a completed snapshot after the fact:
Execution Plan: 01HXM8C2ZK4Q7C2B3F2R4VYV6J Phase: execute Sandbox: git_worktree Worker: local/executor Started: 12:58:10Executing plan [4/4] [x] Collect context [x] Run tools [x] Build changeset [x] Validate
Strategy Summary Decisions: 8 Invariants: 2 Planned Child Plans: 2+ Estimated Files: ~12 Risk: low
[OK] Execution complete — all validations passed
What this looks like in rich format:
The progress indicator is live — the spinner animates, the progress bar fills, and steps transition from pending to active to done in real-time. This snapshot shows the display mid-execution (step 3 active):
╭─ Execution ──────────────────────╮ │ Plan: 01HXM8C2ZK4Q7C2B3F2R4VYV6J │ │ Phase: execute │ │ Sandbox: git_worktree │ │ Worker: local/executor │ │ Started: 12:58:10 │ ╰──────────────────────────────────╯
Executing plan ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 50% elapsed 0:00:13 ✓ Collect context .................. 0.8s ✓ Run tools ...................... 12.4s ⠙ Build changeset ................. (running) ○ Validate ........................ (pending)
When execution completes, the progress indicator resolves to its final state (all steps ✓), the Strategy Summary panel appears below it, and the final status message is displayed.
Example 5: Error Mid-Stream with Partial Output
A command where one of multiple concurrent producers fails, demonstrating graceful partial output.
Producer code (hypothetical agents resource verify):
async def cmd_resource_verify(session: OutputSession, project: str) -> None: """Verify all resources in a project. Some verifications may fail.""" resources = await api.list_project_resources(project)<span style="opacity: 0.7;"># Create a table that will be populated concurrently</span> results_table = session.table(<span style="color: #66cc66;">"Verification Results"</span>, columns=[ ColumnDef(name=<span style="color: #66cc66;">"Resource"</span>, style_hint=<span style="color: #66cc66;">"identifier"</span>), ColumnDef(name=<span style="color: #66cc66;">"Type"</span>), ColumnDef(name=<span style="color: #66cc66;">"Check"</span>), ColumnDef(name=<span style="color: #66cc66;">"Status"</span>), ColumnDef(name=<span style="color: #66cc66;">"Detail"</span>), ]) <span style="opacity: 0.7;"># Progress indicator</span> progress = session.progress( <span style="color: #66cc66;">"Verifying resources"</span>, total=len(resources), steps=[r.name <span style="color: magenta; font-weight: 600;">for</span> r <span style="color: magenta; font-weight: 600;">in</span> resources], ) <span style="opacity: 0.7;"># Verify each resource concurrently</span> <span style="color: magenta; font-weight: 600;">async</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">verify_one</span>(resource): progress.set_step_status(resource.name, <span style="color: #66cc66;">"active"</span>) <span style="color: magenta; font-weight: 600;">try</span>: result = <span style="color: magenta; font-weight: 600;">await</span> api.verify_resource(resource.id) results_table.add_row({ <span style="color: #66cc66;">"Resource"</span>: resource.name, <span style="color: #66cc66;">"Type"</span>: resource.type, <span style="color: #66cc66;">"Check"</span>: result.check_name, <span style="color: #66cc66;">"Status"</span>: <span style="color: #66cc66;">"pass"</span> <span style="color: magenta; font-weight: 600;">if</span> result.passed <span style="color: magenta; font-weight: 600;">else</span> <span style="color: #66cc66;">"fail"</span>, <span style="color: #66cc66;">"Detail"</span>: result.detail, }) progress.set_step_status( resource.name, <span style="color: #66cc66;">"done"</span> <span style="color: magenta; font-weight: 600;">if</span> result.passed <span style="color: magenta; font-weight: 600;">else</span> <span style="color: #66cc66;">"error"</span>, ) <span style="color: magenta; font-weight: 600;">except</span> ApiError <span style="color: magenta; font-weight: 600;">as</span> e: results_table.add_row({ <span style="color: #66cc66;">"Resource"</span>: resource.name, <span style="color: #66cc66;">"Type"</span>: resource.type, <span style="color: #66cc66;">"Check"</span>: <span style="color: #66cc66;">"connection"</span>, <span style="color: #66cc66;">"Status"</span>: <span style="color: #66cc66;">"error"</span>, <span style="color: #66cc66;">"Detail"</span>: <span style="color: cyan;">str</span>(e), }) progress.set_step_status(resource.name, <span style="color: #66cc66;">"error"</span>) progress.increment() <span style="opacity: 0.7;"># Launch all verifications concurrently</span> <span style="color: magenta; font-weight: 600;">await</span> asyncio.gather( *[verify_one(r) <span style="color: magenta; font-weight: 600;">for</span> r <span style="color: magenta; font-weight: 600;">in</span> resources], return_exceptions=<span style="color: magenta; font-weight: 600;">True</span>, <span style="opacity: 0.7;"># Don't fail fast — collect all results</span> ) progress.close() results_table.close() <span style="opacity: 0.7;"># Summarize</span> snapshot = results_table.element pass_count = sum(1 <span style="color: magenta; font-weight: 600;">for</span> r <span style="color: magenta; font-weight: 600;">in</span> snapshot.rows <span style="color: magenta; font-weight: 600;">if</span> r[<span style="color: #66cc66;">"Status"</span>] == <span style="color: #66cc66;">"pass"</span>) fail_count = sum(1 <span style="color: magenta; font-weight: 600;">for</span> r <span style="color: magenta; font-weight: 600;">in</span> snapshot.rows <span style="color: magenta; font-weight: 600;">if</span> r[<span style="color: #66cc66;">"Status"</span>] <span style="color: magenta; font-weight: 600;">in</span> (<span style="color: #66cc66;">"fail"</span>, <span style="color: #66cc66;">"error"</span>)) <span style="color: magenta; font-weight: 600;">if</span> fail_count == 0: session.status(<span style="color: #66cc66;">f"All {pass_count} resources verified"</span>, level=<span style="color: #66cc66;">"ok"</span>) <span style="color: magenta; font-weight: 600;">else</span>: session.status( <span style="color: #66cc66;">f"{fail_count} of {pass_count + fail_count} resources failed verification"</span>, level=<span style="color: #66cc66;">"error"</span>, )
What this produces in plain format (after all concurrent verifications complete):
Verifying resources [3/3] [x] local/api-repo [!] local/staging-db [x] local/docs-repoVerification Results Resource Type Check Status Detail
local/api-repo git-checkout integrity pass All refs valid local/staging-db local/database connection error Connection refused (port 5432) local/docs-repo git-checkout integrity pass All refs valid
[ERROR] 1 of 3 resources failed verification
What this produces in color format:
Verifying resources [3/3] [x] local/api-repo [!] local/staging-db [x] local/docs-repoVerification Results Resource Type Check Status Detail ---------------- -------------- ---------- ------ ---------------------------------- local/api-repo git-checkout integrity pass All refs valid local/staging-db local/database connection error Connection refused (port 5432) local/docs-repo git-checkout integrity pass All refs valid
[ERROR] 1 of 3 resources failed verification
What this produces in yaml format:
command: resource verify
status: error
exit_code: 1
data:
verification_results:
- Resource: local/api-repo
Type: git-checkout
Check: integrity
Status: pass
Detail: All refs valid
- Resource: local/staging-db
Type: local/database
Check: connection
Status: error
Detail: "Connection refused (port 5432)"
- Resource: local/docs-repo
Type: git-checkout
Check: integrity
Status: pass
Detail: All refs valid
timing:
duration_ms: 2840
messages:
- level: error
text: "1 of 3 resources failed verification"
In all formats, the concurrent verification produces a complete result set with partial failures clearly visible. The producer code uses return_exceptions=True on asyncio.gather to ensure all verifications complete even if some fail, and the error handling within each verify_one coroutine ensures that failures are recorded as table rows rather than causing the entire command to abort.
Behavior
Automation Profiles
!!! adr "Architecture Decision" The automation profile system, confidence thresholds, and built-in profiles are defined in ADR-017: Automation Profiles.
Automation profiles determine which phase transitions happen automatically and which require human approval.
Overview
An automation profile is a named collection of confidence thresholds — floating-point values from 0.0 to 1.0 inclusive — that controls which tasks are automated vs. require human approval. Each threshold specifies the minimum confidence score (as computed by the Semantic Escalation system) at which the system proceeds automatically. When the computed confidence for a given operation falls below the profile's threshold for that flag, the system drops to manual mode and requests human input. A threshold of 0.0 means "always automatic" (proceed regardless of confidence), while 1.0 means "always manual" (always require human approval). Profiles follow the same <namespace>/<name> naming convention as actors, tools, skills, and other entities. Profiles are managed via the agents automation-profile CLI commands.
Automatable Tasks
Each automation profile specifies a confidence threshold (a floating-point value from 0.0 to 1.0 inclusive) for each of the following automatable tasks. The threshold determines the minimum confidence level at which the system proceeds automatically. When the computed confidence (from Semantic Escalation) falls below the threshold, the system drops to manual mode for that task. Setting a threshold to 0.0 makes the task always automatic; setting it to 1.0 makes it always manual.
The three safety flags (require_sandbox, require_checkpoints, allow_unsafe_tools) remain boolean since they are binary constraints rather than confidence-dependent behaviors.
| Flag | Type | Description | Behavior |
|---|---|---|---|
auto_strategize |
float (0.0–1.0) | Automatically enter Strategize after plan use |
When confidence >= threshold, Strategize begins immediately. Below threshold, system pauses; user must run agents plan execute to start Strategize. |
auto_execute |
float (0.0–1.0) | Automatically proceed from Strategize to Execute | When confidence >= threshold, Execute begins when Strategize completes. Below threshold, system pauses for user review. |
auto_apply |
float (0.0–1.0) | Automatically proceed from Execute to Apply | When confidence >= threshold, Apply begins when Execute completes. Below threshold, system pauses for user diff review. |
auto_decisions_strategize |
float (0.0–1.0) | Automatically make decisions during Strategize | When confidence >= threshold, the strategy actor makes the decision autonomously. Below threshold, system pauses at the decision point for user input. |
auto_decisions_execute |
float (0.0–1.0) | Automatically make decisions during Execute | When confidence >= threshold, the execution actor makes the decision autonomously. Below threshold, system pauses at the decision point for user input. |
auto_validation_fix |
float (0.0–1.0) | Automatically attempt to fix validation failures | When confidence >= threshold, execution actor self-fixes failing validations. Below threshold, system pauses for user guidance. |
auto_strategy_revision |
float (0.0–1.0) | Automatically revise strategy when Execute hits constraints | When confidence >= threshold, system re-runs Strategize for affected subtree. Below threshold, system pauses and asks user. |
auto_reversion_from_apply |
float (0.0–1.0) | Automatically revert from Apply (constrained state) to Strategize |
When confidence >= threshold, a constrained Apply automatically triggers reversion to Strategize. Below threshold, system pauses and asks user whether to revert or cancel. |
auto_child_plans |
float (0.0–1.0) | Automatically spawn and execute child plans | When confidence >= threshold, child plans are created and executed without pausing. Below threshold, each spawn requires approval. |
auto_retry_transient |
float (0.0–1.0) | Automatically retry on transient failures (network, timeout, rate-limit) | When confidence >= threshold, system retries with backoff. Below threshold, system pauses and asks user. |
auto_checkpoint_restore |
float (0.0–1.0) | Automatically restore from checkpoint on failure | When confidence >= threshold, system rolls back and retries. Below threshold, user decides whether to restore. |
require_sandbox |
boolean | Require sandbox isolation for Execute phase | When true, Execute must run in a sandbox. When false, sandbox is optional. |
require_checkpoints |
boolean | Require checkpointing during Execute | When true, tools must create checkpoints before writes. When false, checkpointing is optional. |
allow_unsafe_tools |
boolean | Allow execution of tools marked as unsafe | When true, unsafe tools can be invoked. When false, unsafe tools are blocked. |
Built-in Automation Profiles
CleverAgents ships with eight built-in automation profiles. Built-in profiles use no namespace prefix.
Threshold values are shown for each flag. A value of 0.0 means always automatic, 1.0 means always manual, and intermediate values (e.g., 0.5, 0.7) mean "automatic when confidence is at or above that level."
| Flag | manual |
review |
supervised |
cautious |
trusted |
auto |
ci |
full-auto |
|---|---|---|---|---|---|---|---|---|
auto_strategize |
1.0 | 0.0 | 0.0 | 0.7 | 0.0 | 0.0 | 0.0 | 0.0 |
auto_execute |
1.0 | 0.0 | 1.0 | 0.7 | 0.0 | 0.0 | 0.0 | 0.0 |
auto_apply |
1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 0.0 | 0.0 |
auto_decisions_strategize |
1.0 | 1.0 | 0.0 | 0.6 | 0.0 | 0.0 | 0.0 | 0.0 |
auto_decisions_execute |
1.0 | 1.0 | 1.0 | 0.8 | 0.0 | 0.0 | 0.0 | 0.0 |
auto_validation_fix |
1.0 | 1.0 | 1.0 | 0.7 | 0.0 | 0.0 | 0.0 | 0.0 |
auto_strategy_revision |
1.0 | 1.0 | 1.0 | 0.8 | 1.0 | 0.0 | 0.0 | 0.0 |
auto_reversion_from_apply |
1.0 | 1.0 | 1.0 | 0.9 | 1.0 | 1.0 | 0.0 | 0.0 |
auto_child_plans |
1.0 | 0.0 | 1.0 | 0.7 | 0.0 | 0.0 | 0.0 | 0.0 |
auto_retry_transient |
1.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 |
auto_checkpoint_restore |
1.0 | 1.0 | 1.0 | 0.6 | 1.0 | 0.0 | 0.0 | 0.0 |
require_sandbox |
true | true | true | true | true | true | true | false |
require_checkpoints |
true | true | true | true | true | true | true | false |
allow_unsafe_tools |
false | false | false | false | false | false | false | true |
manual: Maximum human control. All thresholds set to 1.0 — every phase transition, every decision, every child plan requires explicit human approval regardless of confidence. Sandbox and checkpoints are mandatory. Unsafe tools are blocked. This is the default starting point. Use for: new users learning the system, critical production systems, first-time exploration of an unfamiliar codebase, sensitive projects, regulatory environments.
review: Phases run automatically (Strategize and Execute proceed without pausing), and most decisions within those phases require human approval (thresholds set to 1.0), except transient retries and child plan spawning which are automatic (thresholds 0.0). Apply is manual. The system does the work but consults you on every choice point. Use for: teams that want to stay in the decision loop without manually triggering each phase, code reviews where architectural choices matter more than execution mechanics.
supervised: Strategize thresholds set to 0.0 (always automatic), but Execute and Apply thresholds set to 1.0 (always manual). The system plans autonomously and retries transient failures automatically, but pauses before Execute for human review of the strategy. Use for: projects where you trust the planning but want to review before execution begins.
cautious: Uses intermediate confidence thresholds (0.6–0.8) instead of binary 0.0/1.0 values. The system proceeds automatically only when the Semantic Escalation system reports high confidence, and escalates to the user when uncertain. Apply is always manual (1.0). Higher thresholds (0.8) are applied to riskier operations like strategy revision and execution decisions, while lower thresholds (0.6) are used for safer operations like checkpoint restores and strategy decisions. Transient retries are always automatic. Use for: teams adopting automation gradually, projects with mixed complexity where some tasks are routine but others need oversight, situations where you want the system to self-assess rather than follow rigid rules.
trusted: Strategize and Execute thresholds set to 0.0 (always automatic). Validation fixes, child plan spawning, and transient retries proceed automatically. Strategy revision and checkpoint restores remain manual (thresholds 1.0) to preserve human oversight over plan-level course corrections. The system pauses only before Apply (threshold 1.0) for human review of the final diffs. Use for: day-to-day feature development, routine refactoring, test generation.
auto: All thresholds set to 0.0 except Apply (1.0). The system can revise its own strategy, restore from checkpoints, and handle all decisions automatically. Only Apply requires human approval. Use for: well-understood projects, batch operations, tasks with strong invariant coverage.
ci: All thresholds set to 0.0 — complete end-to-end automation including Apply — but sandbox and checkpoints remain mandatory and unsafe tools are blocked. Designed for non-interactive environments where full automation is needed but safety nets must be preserved. Use for: CI/CD pipelines, automated testing workflows, scheduled batch jobs, any headless execution where rollback capability is essential.
full-auto: All thresholds set to 0.0 — complete end-to-end automation including Apply. No sandbox or checkpoint requirements. Unsafe tools are allowed. Use for: low-risk routine tasks (dependency updates, documentation generation, formatting), trusted batch operations with rollback capabilities, environments where external safety mechanisms exist.
Profile Precedence
Automation profiles are determined using this precedence (highest to lowest):
- Plan-level: Explicitly set via
--automation-profileonagents plan use - Action-level: Set on the action via
--automation-profileonagents action create - Project-level: Set via
agents config set core.automation-profile <PROFILE> --project <PROJECT> - Global-level: Set via
agents config set core.automation-profile <PROFILE>
The effective profile for a plan is resolved at the moment of agents plan use. Once resolved, the profile is locked to that plan — subsequent changes to project or global profiles do not affect running plans.
Child Plan Profile Inheritance
Child plans inherit the parent plan's effective automation profile. If the parent's profile is changed explicitly after creation, new child plans use the new profile while already-running child plans retain their original profile.
Custom Automation Profiles
Custom profiles are created via YAML configuration files and registered with agents automation-profile add. Each automatable task flag takes a confidence threshold (0.0–1.0) instead of a boolean. A threshold of 0.0 means "always automatic" and 1.0 means "always manual." Intermediate values (e.g., 0.5, 0.7) enable fine-grained control where the system proceeds automatically only when the Semantic Escalation confidence score meets or exceeds the threshold:
# File: profiles/careful-auto.yaml name: local/careful-auto description: "Autonomous execution with mandatory sandbox and manual apply"# Confidence thresholds (0.0 = always auto, 1.0 = always manual) auto_strategize: 0.0 auto_execute: 0.0 auto_apply: 1.0 auto_decisions_strategize: 0.0 auto_decisions_execute: 0.0 auto_validation_fix: 0.0 auto_strategy_revision: 1.0 auto_reversion_from_apply: 1.0 auto_child_plans: 0.0 auto_retry_transient: 0.0 auto_checkpoint_restore: 0.0
# Safety flags (boolean) require_sandbox: true require_checkpoints: true allow_unsafe_tools: false
A profile with intermediate thresholds provides nuanced control. For example, auto_decisions_execute: 0.7 means decisions during Execute proceed automatically when confidence >= 0.7, but drop to manual review when confidence is below 0.7:
# File: profiles/nuanced-auto.yaml name: local/nuanced-auto description: "Nuanced automation with graduated confidence thresholds"auto_strategize: 0.0 # Always auto-strategize auto_execute: 0.5 # Auto-execute when confidence >= 0.5 auto_apply: 1.0 # Always require manual apply auto_decisions_strategize: 0.3 # Low bar for strategy decisions auto_decisions_execute: 0.7 # Higher bar for execution decisions auto_validation_fix: 0.5 # Auto-fix when reasonably confident auto_strategy_revision: 0.8 # High confidence needed for auto-revision auto_reversion_from_apply: 0.9 # Very high bar for late-stage reversion auto_child_plans: 0.5 # Auto-spawn when moderately confident auto_retry_transient: 0.0 # Always auto-retry transient errors auto_checkpoint_restore: 0.6 # Auto-restore when fairly confident
require_sandbox: true require_checkpoints: true allow_unsafe_tools: false
agents automation-profile add --config ./profiles/careful-auto.yaml
Semantic Escalation
Semantic Escalation is the system that computes a confidence score (0.0–1.0) for each operation, and it is the mechanism through which automation profile thresholds take effect. The confidence score reflects the system's assessment of how likely an autonomous action is to succeed without human guidance. This score is then compared against the profile's threshold for the relevant flag to determine whether to proceed automatically or escalate to the user.
The confidence score is computed from multiple factors:
class AutonomyController: def should_proceed_automatically(self, decision, context, profile): """Determine whether to proceed automatically or escalate to user.""" factors = { 'past_success_rate': self.get_historical_success(decision.type), 'codebase_familiarity': self.get_familiarity_score(context.project), 'risk_assessment': self.evaluate_risk(decision), 'invariant_complexity': self.analyze_invariants(decision) }confidence = self.compute_confidence(factors) <span style="opacity: 0.7;"># Returns 0.0–1.0</span> threshold = profile.get_threshold(decision.flag) <span style="opacity: 0.7;"># e.g., auto_decisions_execute</span> <span style="color: magenta; font-weight: 600;">if</span> confidence >= threshold: <span style="opacity: 0.7;"># Confidence meets or exceeds the profile threshold — proceed</span> <span style="color: magenta; font-weight: 600;">return</span> ProceedAutonomously(decision, confidence) <span style="color: magenta; font-weight: 600;">else</span>: <span style="opacity: 0.7;"># Confidence below threshold — escalate to human</span> <span style="color: magenta; font-weight: 600;">return</span> RequestHumanGuidance(decision, confidence, threshold, factors)
How thresholds and confidence interact: A profile with auto_decisions_execute: 0.7 means the system will make execution decisions automatically when confidence >= 0.7, but will pause for human input when confidence < 0.7. This is more nuanced than a binary on/off: a profile can set a high bar (0.9) for risky operations while being permissive (0.2) for low-risk ones.
Special cases: A threshold of 0.0 means "always automatic" — even zero confidence passes the threshold. A threshold of 1.0 means "always manual" — no confidence level (which tops out below 1.0 in practice) is high enough to pass. This means the old boolean behavior is a strict subset: true maps to 0.0 and false maps to 1.0.
Progressive Trust Building
New users typically follow this progression:
- Start with
manualto understand system behavior - Move to
reviewto let phases run automatically while staying in the decision loop - Adopt
supervisedas confidence in the planning phase builds - Try
cautiousfor confidence-gated automation that escalates only when uncertain - Adopt
trustedfor routine development tasks - Enable
autofor well-understood projects with strong invariant coverage - Use
cifor headless CI/CD pipelines with safety nets intact - Use
full-autofor low-risk batch operations where external safety mechanisms exist
Guardrails
!!! adr "Architecture Decision" System-level guardrails, pre-flight checks, and runtime limits are defined in ADR-018: Semantic Error Prevention.
System-level guardrails are pre-flight checks and runtime limits that prevent plans from executing in invalid or unsafe conditions. These are distinct from Validations (the Tool subtype described in Core Concepts > Validation) — guardrails operate at the system/infrastructure level before and during plan execution, whereas Validations verify the quality of the work produced by a plan at the end of the Execute phase.
Plan Generation Guardrails
Before a plan begins execution, the system performs pre-flight validation to ensure the plan is well-formed and its dependencies are satisfiable. These checks must be implemented:
- Action schema validation: Verify the action referenced by the plan exists, is well-formed, and its configuration conforms to the expected schema.
- Actor availability: Confirm that all actors required by the plan (strategy actor, execution actor, estimation actor, invariant reconciliation actor) are registered and reachable. For remote actors, verify network connectivity.
- Skill and tool existence: Verify that all skills and tools referenced by the action's actor configuration exist in the registry. This includes both named tools and tools referenced via skills (transitively resolved).
- Automation policy: Verify that the automation profile allows the requested level of autonomy for the target project, resources, and actions.
- Rollback feasibility: If
require_checkpointsis enabled on the plan, verify that all tools in the actor's skill set havecheckpointable: true. Reject plans that would use non-checkpointable tools under a checkpoint-required policy. - Resource accessibility: Verify that the project's linked resources are accessible — git repos are cloneable, databases are connectable, file system paths exist, etc. This is a shallow connectivity check, not a full resource health assessment.
- Validation attachment resolution: Pre-resolve the validations that will apply to this plan (from resource-direct, project, and plan attachment scopes) and verify they are all registered and their tool definitions are valid. This ensures validation failures during Execute are due to the work product, not misconfigured validations.
These pre-flight checks prevent "plan runs with fake providers," missing tool errors mid-execution, and other surprises that waste compute and user time. If any pre-flight check fails, the plan is rejected before entering the Strategize phase, with a clear error message identifying the failing check.
Cost and Rate Limits
Runtime guardrails to prevent runaway resource consumption:
- Per-plan budgets: Maximum API token spend, maximum wall-clock time, maximum number of tool invocations per plan. When a budget is exceeded, the plan pauses and requests user approval to continue or is terminated.
- Per-session budgets: Aggregate limits across all plans in a session. Prevents a single interactive session from consuming excessive resources.
- Per-org budgets: Server-enforced limits for multi-user deployments. Administrators set organization-wide spend caps that cannot be overridden by individual users.
- Per-actor limits: Maximum tool calls per actor invocation, maximum retries per tool failure, maximum tokens per LLM call. Prevents individual actors from looping indefinitely.
- Validation retry limits: Maximum number of fix-then-revalidate iterations when a required Validation fails during Execute. This is a specific application of per-actor limits — the default is 3 attempts, configurable per plan or in the automation profile.
Cost and rate limits are future concerns that require integration with LLM provider billing APIs and internal metering. The system should define configuration surfaces for these limits but may initially implement only the per-plan and per-actor limits, with per-session, per-org, and billing integration added later.
Correcting Plans (Core Feature)
!!! adr "Architecture Decision" Plan correction, decision revision, and the edit-replay model are defined in ADR-007: Decision Tree and Correction.
Correcting plans is where CleverAgents becomes more than "a fancy prompt runner."
The Goal
When a plan makes a wrong decision early, we want to:
- correct the decision,
- recompute only the affected subtree,
- preserve unaffected work.
This is explicitly described: "redo everything below that decision, not the entire code base."
Decision Tree Representation
Every plan records (see Decision Data Model section):
- decisions (choice points) - created during Strategize, including
invariant_enforced,subplan_spawn, andsubplan_parallel_spawndecisions - dependencies (which later work depended on that decision)
- child plans spawned because of that decision - populated during Execute
- artifacts generated under that branch
This makes plan runs auditable and correctable.
Two Correction Modes
- Revert-from-history correction (
--mode=revert)
- Find the decision point in the tree
- Roll back all changes (code and non-code) to that point
- Re-run from that decision point forward
- Keep old execution artifacts for comparison
- Potentially expensive if high up in the tree
- Add-at-end correction (
--mode=append)
- Leave history intact
- Append a new plan at the end that fixes the outcome
- Cheaper and safer sometimes
- Does not rewrite history
Correction Flow (Revert Mode)
When user requests correction at Decision B:
-
Mark for Correction
Decision B.superseded_by = new_decision_id -
Identify Downstream Impact
- Recursively collect all decisions that depend on Decision B (including
invariant_enforced,subplan_spawn, andsubplan_parallel_spawndecisions) - Collect all child plans spawned from those decisions
- These form the "affected subtree"
- Recursively collect all decisions that depend on Decision B (including
-
Rollback Resources
- For each affected decision's
artifacts_produced:- Rollback to the checkpoint before that artifact was created
- For affected child plans:
- Rollback their sandboxes entirely
- For each affected decision's
-
Preserve for Comparison
- Archive the original subtree's artifacts
- Create a
CorrectionAttemptrecord linking old and new
-
Re-execute from Decision Point
- Restore context to
Decision B.context_snapshot - User provides new guidance/correction
- Re-run execution from that point
- New decisions get
is_correction: true, corrects_decision_id: B
- Restore context to
-
Apply as Normal
- The corrected plan goes through normal Apply gating
- Diff shows changes from the correction
History Cleanup
History can only be flagged for cleanup after a plan reaches the applied state.
Once a plan is applied:
- It can no longer be rolled back
- Old correction artifacts can be archived or deleted based on retention policy
- The decision tree is preserved for audit purposes
CLI Commands for Correction
# View decision tree agents plan tree <plan_id> agents --format=json plan tree <plan_id> # For visualization tools# Inspect a specific decision agents plan explain <decision_id> # Shows: question, chosen option, alternatives, rationale, downstream impact
# Correct via revert-and-replay agents plan correct <decision_id> --mode=revert --guidance "<what the decision should be>" # Re-executes from that point with the new guidance
# Correct via append (add fix at end) agents plan correct <decision_id> --mode=append --guidance "<description of the fix>" # Creates a new child plan to fix the outcome without rewriting history
# Compare old vs new after correction agents plan diff --correction <correction_attempt_id>
Correction Safety
Corrections always:
- Create a new attempt revision (increment
plan.attempt) - Preserve old artifacts for diff/compare
- Run execute in sandbox again
- Require apply gating again
- Never modify already-applied changes
This keeps history reproducible and prevents accidental destructive edits.
Human-in-the-Loop Collaboration
!!! adr "Architecture Decision" Human-in-the-loop controls, interruptibility, and approval gates are defined in ADR-017: Automation Profiles.
Even though the direction is "more autonomous," the transcript explicitly recognizes that real workflows require engineers to collaborate with the system, editing code while it works, and using better UX integration (TUI/web/IDE).
So CleverAgents should aim for:
- visibility: what is it doing now?
- interruptibility: pause/cancel/retry
- editability: allow user to modify strategy before execute
- reconciliation: detect if user changed sandbox files mid-run and handle it
UI / Interaction Model
!!! adr "Architecture Decision" The CLI-first interaction model, TUI, and output rendering are defined in ADR-021: CLI and Output Rendering.
CLI-first + TUI + Web App + IDE
The system intends to be CLI-first, with:
- a TUI built using Textual,
- which can generate a web app "for free,"
- and later an IDE plugin that embeds the TUI in the IDE.
This implies a "single UI codebase" model:
- same underlying view logic,
- multiple frontends.
Plan Tree Visualization
The TUI should show:
- plan list
- plan details
- plan tree (ASCII)
- diff view
- approvals
And should later allow exporting the tree as image (PNG) or JSON for other visualization tools.
Configuration
!!! adr "Architecture Decision" The configuration system design principles, YAML-first approach, and config file conventions are defined in ADR-024: Configuration System.
This section provides a complete reference for the YAML configuration files used to define the major configurable objects in CleverAgents: Actors, Skills, Tools, Actions, Resource Types, Context Views, and Automation Profiles. Projects are created via CLI commands rather than standalone configuration files, but their context views are configured through YAML and are documented here as well. It also documents the global configuration keys that control system-wide behavior.
Global Configuration Keys
!!! adr "Architecture Decision" Global configuration keys, hierarchical key structure, and config precedence are defined in ADR-024: Configuration System.
Global configuration keys control system-wide defaults and behavior. They are stored in the global configuration file (default: ~/.cleveragents/config.toml) and managed via the agents config set, agents config get, and agents config list commands.
Hierarchical Key Structure
Configuration keys use dot-separated hierarchical names, similar to git config. Each dot introduces a level of nesting, grouping related keys under a common parent. For example:
core.format # top-level "core" group, "format" key
core.log.level # "core" group, "log" subgroup, "level" key
index.text.backend # "index" group, "text" subgroup, "backend" key
The key plan.concurrency is not the same as a top-level key concurrency — the dot path is the full identity of the key. This means you can have plan.concurrency and session.concurrency as distinct keys without conflict.
On the CLI, the full dot path is always used:
$ agents config set core.format table
$ agents config get plan.budget.per-plan
$ agents config list plan.*
The agents config list command accepts an optional glob pattern to filter keys by subtree. For example, agents config list index.* lists all keys under the index group, and agents config list provider.* lists all provider credential keys.
In the TOML configuration file, the hierarchy maps directly to TOML's native table structure. The dot-path key index.text.backend can be expressed equivalently as:
# Inline dot notation (convenient for single keys) index.text.backend = "tantivy"# Or as nested TOML tables (better for groups) [index.text] backend = "tantivy"
[index.vector] backend = "faiss"
Both forms are valid and can be mixed. The TOML parser normalizes them to the same internal representation.
Every key follows a resolution chain: CLI flag > environment variable > project-scoped config > global config file > built-in default.
Key Reference
Keys are organized by their top-level group. Within each group, the full dot-path, type, default, corresponding environment variable, and description are provided.
core.* — Core System Settings
| Key | Type | Default | Env Variable | Description |
|---|---|---|---|---|
core.data-dir |
string | ~/.cleveragents |
CLEVERAGENTS_DATA_DIR |
Root directory for all CleverAgents persistent state: database, logs, cache, sessions, checkpoints, and backups. All other paths are derived relative to this directory unless individually overridden. Overridden per-invocation by the --data-dir CLI flag. |
core.config-path |
string | <core.data-dir>/config.toml |
CLEVERAGENTS_CONFIG_PATH |
Path to the global configuration file itself. Typically only set via environment variable or CLI flag to bootstrap an alternate configuration. Overridden per-invocation by the --config-path CLI flag. |
core.format |
string | rich |
CLEVERAGENTS_FORMAT |
Default output rendering format for all CLI commands. Accepted values: rich, color, table, plain, json, yaml. Overridden per-invocation by the --format CLI flag. |
core.namespace |
string | local |
CLEVERAGENTS_NAMESPACE |
The default namespace prefix applied when creating entities without an explicit namespace. For local-only usage, this is local. In server mode, typically the user's username or organization name. |
core.automation-profile |
string | supervised |
CLEVERAGENTS_AUTOMATION_PROFILE |
The default automation profile applied to new plans when no profile is specified at the plan, action, or project level. Accepts any built-in profile name (manual, review, supervised, cautious, trusted, auto, ci, full-auto) or a custom profile name in <namespace>/<name> format. Project-scopable. |
core.log.level |
string | FATAL |
CLEVERAGENTS_LOG_LEVEL |
Base logging verbosity level. Accepted values (from least to most verbose): FATAL, ERROR, WARN, INFO, DEBUG, TRACE. The default FATAL means only fatal errors (where the application exits) produce log output; all other log messages are suppressed. The -v CLI flag can be repeated (-v through -vvvvv) to raise the effective level for a single invocation: -v = ERROR, -vv = WARN, -vvv = INFO, -vvvv = DEBUG, -vvvvv = TRACE. The effective level applies to both log file output and terminal log output (subject to core.log.terminal and core.log.file-enabled settings). |
core.log.dir |
string | <core.data-dir>/logs |
CLEVERAGENTS_LOG_DIR |
Directory where log files are written. Can be set to an alternative path to redirect log storage. |
core.log.file-enabled |
boolean | true |
CLEVERAGENTS_LOG_FILE_ENABLED |
Whether log output is written to log files. When true, log messages at or above the effective verbosity level are written to the log directory. When false, log file output is suppressed entirely (useful when log output should only go to the terminal). |
core.log.terminal |
string | auto |
CLEVERAGENTS_LOG_TERMINAL |
Controls whether log output raised by -v is also displayed on the terminal. Accepted values: auto (show on terminal when -v is used, suppress otherwise), always (always show log output on terminal at the effective level), never (suppress terminal log output entirely — log only to file). This keeps log output separate from normal program output, which goes to stdout. |
core.log.terminal-stream |
string | stderr |
CLEVERAGENTS_LOG_TERMINAL_STREAM |
The stream used for terminal log output. Accepted values: stderr (default, keeps logs separate from normal stdout output), stdout (merge with normal output), or a file path / device path (e.g., /dev/null, /dev/pts/1, or a file path) to redirect terminal log output to an alternative destination. |
core.log.retention-days |
integer | 30 |
CLEVERAGENTS_LOG_RETENTION_DAYS |
Number of days to retain log files before automatic cleanup. |
core.backup.dir |
string | <core.data-dir>/backups |
CLEVERAGENTS_BACKUP_DIR |
Directory where backup snapshots are stored. |
core.backup.retention-days |
integer | 7 |
CLEVERAGENTS_BACKUP_RETENTION_DAYS |
Number of days to retain backup snapshots created during project deletion, agents init resets, and plan correction history cleanup. Backups older than this are automatically purged. |
core.cache.dir |
string | <core.data-dir>/cache |
CLEVERAGENTS_CACHE_DIR |
Directory for transient caches (downloaded models, tool artifacts, temporary files). |
server.* — Server Mode
| Key | Type | Default | Env Variable | Description |
|---|---|---|---|---|
server.url |
string | (not set) | CLEVERAGENTS_SERVER_URL |
URL of the CleverAgents server for multi-user collaborative mode. When set, the CLI operates in server mode, syncing namespaced entities (actors, actions, skills, projects) with the remote server. When unset, the CLI operates in local-only mode. |
server.token |
string | (not set) | CLEVERAGENTS_SERVER_TOKEN |
Authentication token for server mode. Required when server.url is set. Obtained via server registration or team invite flow. Prefer the environment variable for CI environments. |
server.sync.auto |
boolean | true |
CLEVERAGENTS_SERVER_SYNC_AUTO |
Whether to automatically sync entity definitions (actors, actions, skills) with the server on startup and after registration changes. When false, sync must be triggered manually. |
server.sync.interval |
integer | 300 |
CLEVERAGENTS_SERVER_SYNC_INTERVAL |
Interval in seconds between automatic background syncs with the server. Only applies when server.sync.auto is true. |
actor.* — Actor Defaults
| Key | Type | Default | Env Variable | Description |
|---|---|---|---|---|
actor.default.strategy |
string | (not set) | CLEVERAGENTS_DEFAULT_STRATEGY_ACTOR |
Default strategy actor used for plans when not specified on the action or plan use invocation. Must reference a registered actor in <namespace>/<name> format. When unset, the action's strategy_actor field is required. |
actor.default.execution |
string | (not set) | CLEVERAGENTS_DEFAULT_EXECUTION_ACTOR |
Default execution actor used for plans when not specified on the action or plan use invocation. Must reference a registered actor. When unset, the action's execution_actor field is required. |
actor.default.estimation |
string | (not set) | CLEVERAGENTS_DEFAULT_ESTIMATION_ACTOR |
Default estimation actor used for cost and effort estimation. Estimation is optional; when unset, plans skip the estimation step. |
actor.default.invariant |
string | (not set) | CLEVERAGENTS_DEFAULT_INVARIANT_ACTOR |
Default Invariant Reconciliation Actor used globally. When a plan enters Strategize and neither the plan, action, nor project specifies an invariant actor, this actor reconciles invariant conflicts across scopes. Must reference a registered actor. |
actor.default.orchestrator |
string | (not set) | CLEVERAGENTS_DEFAULT_ORCHESTRATOR |
Default orchestrator actor for new sessions created without an explicit --actor flag. |
plan.* — Plan Execution
| Key | Type | Default | Env Variable | Description |
|---|---|---|---|---|
plan.concurrency |
integer | 4 |
CLEVERAGENTS_PLAN_CONCURRENCY |
Maximum number of plans that can execute concurrently. When this limit is reached, new plan execute commands queue until a slot opens. Child plans count toward the parent's concurrency allocation, not toward the global limit. Project-scopable. |
plan.max-child-depth |
integer | 5 |
CLEVERAGENTS_PLAN_MAX_CHILD_DEPTH |
Maximum nesting depth for child plan spawning. Prevents runaway recursive plan decomposition. A value of 1 means no child plans; 5 means up to 5 levels of parent-child nesting. Project-scopable. |
plan.budget.per-plan |
float | (not set) | CLEVERAGENTS_PLAN_BUDGET |
Maximum estimated API cost (in USD) allowed per plan. When the running cost estimate exceeds this budget, the plan pauses and requests human approval to continue. When unset, no per-plan cost limit is enforced. Project-scopable. |
plan.budget.per-session |
float | (not set) | CLEVERAGENTS_SESSION_BUDGET |
Maximum estimated API cost (in USD) allowed per session. When cumulative session cost exceeds this budget, all plan operations in the session pause. When unset, no per-session cost limit is enforced. |
plan.budget.warn-threshold |
float | 0.8 |
CLEVERAGENTS_PLAN_BUDGET_WARN |
Fraction of the per-plan budget at which a warning is emitted. For example, 0.8 means a warning fires at 80% of the budget. Only effective when plan.budget.per-plan is set. |
plan.tool.max-calls-per-step |
integer | 25 |
CLEVERAGENTS_TOOL_MAX_CALLS |
Maximum number of tool invocations allowed in a single actor step (one LLM turn). Prevents runaway tool loops. When the limit is reached, the actor step is interrupted and the plan pauses for human review. |
plan.tool.max-retries |
integer | 3 |
CLEVERAGENTS_TOOL_MAX_RETRIES |
Maximum number of retry attempts for a single failing tool invocation before escalating. Applies to transient errors (network timeouts, rate limits). Non-transient errors fail immediately. |
plan.tool.retry-backoff |
string | exponential |
CLEVERAGENTS_TOOL_RETRY_BACKOFF |
Retry backoff strategy for transient tool failures. Accepted values: exponential (doubling delay starting at 1s), linear (fixed 2s delay), none (immediate retry). |
sandbox.* — Sandbox and Checkpointing
| Key | Type | Default | Env Variable | Description |
|---|---|---|---|---|
sandbox.strategy |
string | git_worktree |
CLEVERAGENTS_SANDBOX_STRATEGY |
Default sandbox isolation strategy for the Execute phase. Accepted values: git_worktree (create a git worktree for isolation), filesystem_copy (copy the project directory), transaction_rollback (for database resources), none (no isolation — requires require_sandbox: false in the automation profile). Per-resource-type strategies override this default. Project-scopable. |
sandbox.cleanup |
string | on_apply |
CLEVERAGENTS_SANDBOX_CLEANUP |
When to clean up sandbox working directories. Accepted values: on_apply (clean up after successful apply), on_terminal (clean up when plan reaches any terminal state — applied, cancelled, or failed), manual (never auto-clean; user must delete). |
sandbox.checkpoint.enabled |
boolean | true |
CLEVERAGENTS_CHECKPOINT_ENABLED |
Whether checkpointing is enabled globally. When true, the Execute phase creates checkpoints before write operations, enabling rollback. When false, checkpointing is skipped (requires require_checkpoints: false in the automation profile). Project-scopable. |
sandbox.checkpoint.dir |
string | <core.data-dir>/checkpoints |
CLEVERAGENTS_CHECKPOINT_DIR |
Directory where plan execution checkpoints are stored. Each plan gets a subdirectory named by its ULID. |
sandbox.checkpoint.max-per-plan |
integer | 50 |
CLEVERAGENTS_CHECKPOINT_MAX |
Maximum number of checkpoints retained per plan. When exceeded, the oldest checkpoints are pruned (keeping the first and most recent). |
index.* — Code Intelligence and Indexing
| Key | Type | Default | Env Variable | Description |
|---|---|---|---|---|
index.text.backend |
string | tantivy |
CLEVERAGENTS_INDEX_TEXT_BACKEND |
Backend for full-text search indexing of project resources. Accepted values: tantivy (recommended, high-performance Rust-based engine), sqlite_fts (built-in SQLite FTS5, no additional dependencies). |
index.text.dir |
string | <core.data-dir>/index/text |
CLEVERAGENTS_INDEX_TEXT_DIR |
Directory where the full-text search index is stored. |
index.vector.backend |
string | faiss |
CLEVERAGENTS_INDEX_VECTOR_BACKEND |
Backend for vector similarity search (semantic code search). Accepted values: faiss (Facebook AI Similarity Search, CPU or GPU), qdrant (requires a running Qdrant server), none (disable vector search). |
index.vector.dir |
string | <core.data-dir>/index/vector |
CLEVERAGENTS_INDEX_VECTOR_DIR |
Directory where the vector index is stored (for local backends). |
index.vector.qdrant-url |
string | (not set) | CLEVERAGENTS_QDRANT_URL |
URL of the Qdrant server. Required when index.vector.backend is qdrant. |
index.graph.backend |
string | none |
CLEVERAGENTS_INDEX_GRAPH_BACKEND |
Backend for the knowledge graph store (structural code relationships, RDF triples). Accepted values: neo4j (requires a running Neo4j server), rdflib (in-process Python RDF library), none (disable graph queries). |
index.graph.neo4j-url |
string | (not set) | CLEVERAGENTS_NEO4J_URL |
URL of the Neo4j server. Required when index.graph.backend is neo4j. |
index.graph.neo4j-auth |
string | (not set) | CLEVERAGENTS_NEO4J_AUTH |
Authentication for Neo4j in user:password format. Required when index.graph.backend is neo4j. |
index.embedding.provider |
string | openai |
CLEVERAGENTS_EMBEDDING_PROVIDER |
Provider used for generating vector embeddings. Accepted values: openai, anthropic, local (uses a local sentence-transformers model). When local, no API key is required for embedding generation. |
index.embedding.model |
string | text-embedding-3-small |
CLEVERAGENTS_EMBEDDING_MODEL |
Model identifier used for generating vector embeddings when indexing project resources. Must be a valid model for the configured index.embedding.provider. |
index.embedding.dimensions |
integer | (provider default) | CLEVERAGENTS_EMBEDDING_DIMENSIONS |
Number of dimensions for generated embeddings. When unset, the provider's default dimensionality for the chosen model is used. Can be reduced for faster search at the cost of accuracy. |
index.auto-reindex |
boolean | true |
CLEVERAGENTS_AUTO_REINDEX |
Whether to automatically re-index project resources when files change on disk. When true, indexes are updated on resource add and before each Strategize phase. When false, re-indexing must be triggered manually. |
context.* — Context Tier Defaults
These keys set the default context policy for all projects. Project-level context policies (set via agents project context set) override these defaults. All keys in this group are project-scopable.
| Key | Type | Default | Env Variable | Description |
|---|---|---|---|---|
context.hot.max-tokens |
integer | 16000 |
CLEVERAGENTS_CTX_HOT_TOKENS |
Default maximum token budget for hot context (the current working set loaded into the LLM context window). This is a soft cap; the actor's model context window is the hard limit. |
context.warm.max-decisions |
integer | 100 |
CLEVERAGENTS_CTX_WARM_DECISIONS |
Default maximum number of recent decisions retained in warm context (available for retrieval but not loaded by default). |
context.cold.max-decisions |
integer | 500 |
CLEVERAGENTS_CTX_COLD_DECISIONS |
Default maximum number of historical decisions retained in cold context (archived for audit and correction). |
context.query.limit |
integer | 20 |
CLEVERAGENTS_CTX_QUERY_LIMIT |
Default maximum number of retrieval results returned per context query during Strategize and Execute. |
context.query.min-relevance |
float | 0.3 |
CLEVERAGENTS_CTX_QUERY_MIN_RELEVANCE |
Minimum relevance score (0.0–1.0) for retrieval results to be included in context. Results below this threshold are discarded even if the query limit is not reached. |
context.file.max-size |
integer | 1048576 |
CLEVERAGENTS_CTX_MAX_FILE_SIZE |
Default maximum file size (in bytes) for files included in context. Files exceeding this size are summarized or excluded. Default is 1 MB (1,048,576 bytes). |
context.file.max-total-size |
integer | 52428800 |
CLEVERAGENTS_CTX_MAX_TOTAL_SIZE |
Default maximum total size (in bytes) across all files included in context. Default is 50 MB. |
context.summarize.enabled |
boolean | true |
CLEVERAGENTS_CTX_SUMMARIZE |
Whether to summarize large context segments that exceed token limits rather than truncating them. When true, an LLM summarization pass produces a condensed version. When false, content is hard-truncated. |
context.summarize.max-tokens |
integer | 1000 |
CLEVERAGENTS_CTX_SUMMARY_TOKENS |
Maximum number of tokens in a generated context summary. Controls the verbosity of summaries produced when context.summarize.enabled is true. |
context.summarize.model |
string | (uses actor's model) | CLEVERAGENTS_CTX_SUMMARY_MODEL |
Model to use for context summarization. When unset, the active actor's own model is used. Set this to a faster/cheaper model (e.g., a smaller variant) to reduce summarization cost. |
context.strategies.enabled |
list | ["simple-keyword", "semantic-embedding", "breadth-depth-navigator"] |
CLEVERAGENTS_CTX_STRATEGIES |
Ordered list of ACMS context strategies enabled globally. Strategies are run in parallel and their results fused by the Context Assembly Pipeline. |
context.strategies.custom.* |
object | (none) | — | Registration block for custom context strategies. Each key under custom names a strategy; the value specifies its module, class, max_quality, and optional config. |
context.strategies.arce.model |
string | (uses actor's model) | CLEVERAGENTS_CTX_ARCE_MODEL |
Model used by the ARCE (Autonomous Reasoning Context Extraction) strategy for its internal reasoning loop. |
context.strategies.arce.max-rounds |
integer | 3 |
CLEVERAGENTS_CTX_ARCE_ROUNDS |
Maximum number of search-refine rounds the ARCE strategy performs before returning results. |
context.strategies.breadth-depth-navigator.max-hops |
integer | 4 |
CLEVERAGENTS_CTX_BDN_HOPS |
Maximum graph traversal depth for the breadth-depth-navigator strategy. |
context.budget.response-reserve-tokens |
integer | 4096 |
CLEVERAGENTS_CTX_RESPONSE_RESERVE |
Number of tokens reserved from the model's context window for the response. Subtracted before computing the context budget. |
context.budget.tool-definition-estimate |
integer | (auto-computed) | CLEVERAGENTS_CTX_TOOL_ESTIMATE |
Estimated tokens consumed by tool definitions in the prompt. When unset, computed from the active skill set. |
context.budget.skeleton-ratio |
float | 0.15 |
CLEVERAGENTS_CTX_SKELETON_RATIO |
Default fraction of the context budget reserved for inherited plan skeleton context. |
context.budget.refresh-threshold |
float | 0.30 |
CLEVERAGENTS_CTX_REFRESH_THRESHOLD |
Fractional change in available budget that triggers automatic context re-assembly. E.g., 0.30 means re-assemble when budget changes by more than 30%. |
context.budget.min-useful-budget |
integer | 500 |
CLEVERAGENTS_CTX_MIN_BUDGET |
Minimum useful context budget in tokens. If the computed budget falls below this, context assembly is skipped and a warning is emitted. |
context.tiers.warm.retention-hours |
integer | 24 |
CLEVERAGENTS_CTX_WARM_HOURS |
Number of hours warm-tier context entries are retained before demotion to cold tier. |
context.tiers.cold.retention-days |
integer | 90 |
CLEVERAGENTS_CTX_COLD_DAYS |
Number of days cold-tier context entries are retained before archival/expiry. |
context.uko.default-analyzers |
list | ["python", "typescript", "rust", "java", "markdown", "json-schema"] |
CLEVERAGENTS_CTX_UKO_ANALYZERS |
List of UKO analyzers enabled by default for resource indexing. Each analyzer extracts language-specific UKO nodes from resources. |
context.uko.analyzers.custom.* |
object | (none) | — | Registration block for custom UKO analyzers. Each key under custom names an analyzer; the value specifies its module, class, languages, and optional config. |
context.pipeline.strategy-selector |
string | "builtin:ConfidenceWeightedSelector" |
CLEVERAGENTS_CTX_PIPELINE_SELECTOR |
Implementation class for the StrategySelector pipeline component. Decides which strategies to invoke and with what confidence. Value is "module:ClassName". |
context.pipeline.budget-allocator |
string | "builtin:ProportionalBudgetAllocator" |
CLEVERAGENTS_CTX_PIPELINE_ALLOCATOR |
Implementation class for the BudgetAllocator pipeline component. Distributes token budget across selected strategies. |
context.pipeline.strategy-executor |
string | "builtin:ParallelStrategyExecutor" |
CLEVERAGENTS_CTX_PIPELINE_EXECUTOR |
Implementation class for the StrategyExecutor pipeline component. Controls parallelism, timeouts, and circuit breaking for strategy invocation. |
context.pipeline.fragment-deduplicator |
string | "builtin:ContentHashDeduplicator" |
CLEVERAGENTS_CTX_PIPELINE_DEDUP |
Implementation class for the FragmentDeduplicator pipeline component. Removes duplicate fragments via content-hash, UKO-identity, or semantic similarity. |
context.pipeline.detail-depth-resolver |
string | "builtin:MaxDepthResolver" |
CLEVERAGENTS_CTX_PIPELINE_DEPTH |
Implementation class for the DetailDepthResolver pipeline component. Resolves conflicts when the same UKO node appears at different detail depths. |
context.pipeline.fragment-scorer |
string | "builtin:WeightedCompositeScorer" |
CLEVERAGENTS_CTX_PIPELINE_SCORER |
Implementation class for the FragmentScorer pipeline component. Computes composite relevance scores for ranking during budget packing. |
context.pipeline.budget-packer |
string | "builtin:GreedyKnapsackPacker" |
CLEVERAGENTS_CTX_PIPELINE_PACKER |
Implementation class for the BudgetPacker pipeline component. Fits scored fragments into the token budget with depth fallback support. |
context.pipeline.fragment-orderer |
string | "builtin:RelevanceCoherenceOrderer" |
CLEVERAGENTS_CTX_PIPELINE_ORDERER |
Implementation class for the FragmentOrderer pipeline component. Orders packed fragments for optimal coherence in the context window. |
context.pipeline.preamble-generator |
string | "builtin:ProvenancePreambleGenerator" |
CLEVERAGENTS_CTX_PIPELINE_PREAMBLE |
Implementation class for the PreambleGenerator pipeline component. Generates provenance summaries prepended to assembled context. |
context.pipeline.skeleton-compressor |
string | "builtin:DepthReductionCompressor" |
CLEVERAGENTS_CTX_PIPELINE_COMPRESSOR |
Implementation class for the SkeletonCompressor pipeline component. Compresses parent context into skeleton for child plan inheritance. |
context.pipeline.strategy-executor.timeout-seconds |
integer | 30 |
CLEVERAGENTS_CTX_PIPELINE_EXEC_TIMEOUT |
Timeout in seconds for each strategy execution within the StrategyExecutor. |
context.pipeline.strategy-executor.max-workers |
integer | 4 |
CLEVERAGENTS_CTX_PIPELINE_EXEC_WORKERS |
Maximum number of parallel workers for concurrent strategy execution. |
context.pipeline.strategy-executor.circuit-breaker-threshold |
integer | 3 |
CLEVERAGENTS_CTX_PIPELINE_EXEC_CB |
Number of consecutive failures before the circuit breaker opens for a strategy. |
context.pipeline.fragment-scorer.relevance-weight |
float | 0.4 |
CLEVERAGENTS_CTX_PIPELINE_SCORER_REL |
Weight of relevance score in composite fragment scoring. |
context.pipeline.fragment-scorer.hierarchy-weight |
float | 0.3 |
CLEVERAGENTS_CTX_PIPELINE_SCORER_HIER |
Weight of hierarchy position in composite fragment scoring. |
context.pipeline.fragment-scorer.quality-weight |
float | 0.2 |
CLEVERAGENTS_CTX_PIPELINE_SCORER_QUAL |
Weight of strategy quality in composite fragment scoring. |
context.pipeline.fragment-scorer.recency-weight |
float | 0.1 |
CLEVERAGENTS_CTX_PIPELINE_SCORER_REC |
Weight of recency bonus in composite fragment scoring. |
context.pipeline.budget-packer.depth-fallback-steps |
list | [9, 4, 2, 0] |
CLEVERAGENTS_CTX_PIPELINE_PACKER_STEPS |
Ordered list of detail depths to try during depth fallback when a fragment doesn't fit at its current depth. |
context.pipeline.budget-packer.min-fragment-tokens |
integer | 10 |
CLEVERAGENTS_CTX_PIPELINE_PACKER_MIN |
Minimum token size for a fragment to be included. Fragments below this threshold are excluded. |
context.pipeline.preamble-generator.enabled |
boolean | true |
CLEVERAGENTS_CTX_PIPELINE_PREAMBLE_ON |
Whether the PreambleGenerator produces a preamble. When false, no preamble is prepended to assembled context. |
context.pipeline.preamble-generator.max-tokens |
integer | 200 |
CLEVERAGENTS_CTX_PIPELINE_PREAMBLE_TOK |
Maximum tokens allocated to the context preamble. |
provider.* — LLM Provider Credentials
Provider credential keys follow the pattern provider.<name>.<field>. Each provider has an api-key field and may have additional fields for endpoint configuration. Environment variables for provider keys use standard naming conventions compatible with existing tooling (e.g., OPENAI_API_KEY).
| Key | Type | Default | Env Variable | Description |
|---|---|---|---|---|
provider.openai.api-key |
string | (not set) | OPENAI_API_KEY |
API key for OpenAI models. Required when using OpenAI-based actors or the default embedding provider. The standard OPENAI_API_KEY environment variable is checked for compatibility with existing tooling. |
provider.openai.org-id |
string | (not set) | OPENAI_ORG_ID |
OpenAI organization ID. Optional; used when the API key belongs to multiple organizations. |
provider.openai.base-url |
string | (not set) | OPENAI_BASE_URL |
Custom base URL for OpenAI-compatible APIs (e.g., a local proxy or alternative provider). When unset, the official OpenAI API URL is used. |
provider.anthropic.api-key |
string | (not set) | ANTHROPIC_API_KEY |
API key for Anthropic models. Required when using Anthropic-based actors. |
provider.google.api-key |
string | (not set) | GOOGLE_API_KEY |
API key for Google AI models. Required when using Google-based actors. |
provider.azure.endpoint |
string | (not set) | AZURE_OPENAI_ENDPOINT |
Azure OpenAI endpoint URL. Required when using Azure-hosted OpenAI models. |
provider.azure.api-key |
string | (not set) | AZURE_OPENAI_API_KEY |
API key for Azure OpenAI. Required alongside provider.azure.endpoint. |
provider.azure.api-version |
string | 2024-02-01 |
AZURE_OPENAI_API_VERSION |
Azure OpenAI API version string. |
provider.openrouter.api-key |
string | (not set) | OPENROUTER_API_KEY |
API key for OpenRouter. Required when using OpenRouter as a provider. |
Configuration Scoping
Several keys support project-scoped values in addition to the global default. When a project-scoped value is set, it applies only to plans targeting that project. Keys marked Project-scopable in the reference above support this feature.
The full dot path is used identically in both global and project-scoped contexts:
# Set a global default $ agents config set core.automation-profile trusted# Override for a specific project — same key, scoped to the project $ agents config set core.automation-profile manual --project local/production-api
# Project-scoped keys under any group work the same way $ agents config set plan.budget.per-plan 2.00 --project local/production-api $ agents config set sandbox.strategy git_worktree --project local/production-api $ agents config set context.hot.max-tokens 32000 --project local/large-codebase
The full list of project-scopable keys: core.automation-profile, plan.concurrency, plan.max-child-depth, plan.budget.per-plan, sandbox.strategy, sandbox.checkpoint.enabled, and all keys under context.*.
Resolution Order
When resolving the effective value of a configuration key, the following precedence applies (highest to lowest):
- CLI flag: Per-invocation overrides (e.g.,
--formatmaps tocore.format,--data-dirmaps tocore.data-dir). Highest priority. - Environment variable: The mapped
CLEVERAGENTS_*variable or provider-specific variable (e.g.,OPENAI_API_KEYforprovider.openai.api-key). - Project-scoped config: For project-scopable keys, the value set via
agents config set <key> <value> --project <PROJECT>. Only applies in contexts where a project is known (e.g., during plan execution). - Global config file: The value in the config file at the path specified by
core.config-path. - Built-in default: The hardcoded default value.
The agents config get <key> command displays the full resolution chain, showing which level provided the winning value. Listing subtrees is supported:
# Show the full resolution chain for a single key $ agents config get sandbox.checkpoint.enabled# List all keys in a group $ agents config list index.*
# List all keys with their current effective values $ agents config list
Configuration File Format
The global configuration file uses TOML format. The hierarchical key structure maps naturally to TOML tables. A typical config.toml after initial setup:
# ~/.cleveragents/config.toml[core] data-dir = "/home/alex/.cleveragents" format = "rich" namespace = "local" automation-profile = "supervised"
[core.log] level = "FATAL" terminal = "auto" terminal-stream = "stderr" file-enabled = true retention-days = 30
[core.backup] retention-days = 7
[server] # url = "https://agents.example.com" # token = "tok_01HXR..."
[actor.default] strategy = "local/strategist" execution = "local/executor" invariant = "local/invariant-resolver" # estimation = "local/estimator" # orchestrator = "local/orchestrator"
[plan] concurrency = 4 max-child-depth = 5
[plan.budget] per-plan = 5.00 per-session = 25.00 warn-threshold = 0.8
[plan.tool] max-calls-per-step = 25 max-retries = 3 retry-backoff = "exponential"
[sandbox] strategy = "git_worktree" cleanup = "on_apply"
[sandbox.checkpoint] enabled = true max-per-plan = 50
[index.text] backend = "tantivy"
[index.vector] backend = "faiss"
[index.graph] backend = "none"
[index.embedding] provider = "openai" model = "text-embedding-3-small"
[context.hot] max-tokens = 16000
[context.warm] max-decisions = 100
[context.cold] max-decisions = 500
[context.query] limit = 20 min-relevance = 0.3
[context.file] max-size = 1048576 max-total-size = 52428800
[context.summarize] enabled = true max-tokens = 1000
# Provider keys — prefer environment variables for secrets [provider.openai] # api-key = "sk-..."
[provider.anthropic] # api-key = "sk-ant-..."
# Project-scoped overrides [project."local/production-api"] "core.automation-profile" = "manual" "plan.budget.per-plan" = 2.00 "sandbox.strategy" = "git_worktree" "sandbox.checkpoint.enabled" = true
[project."local/docs"] "core.automation-profile" = "auto" "plan.budget.per-plan" = 10.00 "context.hot.max-tokens" = 32000
Design Principles
All CleverAgents configuration files share a consistent set of design principles:
-
YAML-first: All configuration files use YAML syntax. JSON is also accepted where a configuration file is loaded (since valid JSON is valid YAML), but YAML is the canonical format for human-authored configurations.
-
Namespace/Name convention: Every configurable entity follows the
<namespace>/<name>naming convention. Thelocal/namespace is reserved for local-only items. User namespaces (<username>/) and organization namespaces (<orgname>/) are stored on the server. Built-in entities use provider namespaces (e.g.,openai/,anthropic/). -
Config-as-complete-definition pattern: For entity registration commands (
actor add,skill add,tool add,validation add,resource type add,action create,automation-profile add), the YAML configuration file is the sole source of truth — the--configfile fully defines the entity and no CLI override flags are accepted. For runtime commands that reference existing entities (e.g.,plan use), CLI options may override entity defaults to customize behavior per-invocation. -
Environment variable interpolation: All configuration files support
${ENV_VAR}and${ENV_VAR:default_value}syntax for environment variable interpolation. If a variable is not set and no default is provided, an error is raised. Boolean strings (true/false) and numeric strings are automatically converted to their native types. -
Jinja2 template support: Actor configuration files support Jinja2 template syntax within string values (particularly
system_promptfields). Template placeholders like{{ context.project_name }}are preserved during YAML parsing and rendered at runtime. -
Idempotent registration: Adding an entity that already exists fails unless
--updateis provided. This prevents accidental overwrites while allowing intentional updates. -
Declarative structure: Configuration files are declarative — they describe what should exist, not how to create it. The system handles all lifecycle management.
-
Schema version: Configuration files may include a top-level
cleveragents.versionfield to indicate which schema version they target (currently"3.0"). When omitted, the latest schema version is assumed.
Actor Configuration Files
!!! adr "Architecture Decision" The actor configuration schema and actor registration model are defined in ADR-010: Actor and Agent Architecture.
Actor configuration files define intelligent agents — anything conversational, from a single LLM to an entire graph of interconnected actors and tools. Actors are registered via agents actor add --config <file>.
JSON Schema
The following is the formal JSON Schema definition for actor configuration files. Since YAML is a superset of JSON, this schema can be used directly to validate actor YAML files using any JSON Schema validator.
{
"$schema": "https://json-schema.org/draft/2020-12/schema",
"$id": "https://cleveragents.dev/schemas/actor-config.json",
"title": "CleverAgents Actor Configuration",
"description": "Configuration file schema for defining CleverAgents actors — intelligent agents composed of LLMs, tools, and graph topologies.",
"type": "object",
"properties": {
"name": {
"type": "string",
"pattern": "^[a-z0-9_-]+/[a-z0-9_-]+$",
"description": "Namespaced actor name in <namespace>/<name> format (e.g., 'local/reviewer', 'myorg/deploy-agent'). Required. When the actor is added via the CLI, this value is used as the registered name."
},
"cleveragents": {
"type": "object",
"description": "Metadata block containing schema version, logging, template engine, and safety settings.",
"properties": {
"version": {
"type": "string",
"description": "Schema version for this configuration file.",
"default": "3.0"
},
"logging": {
"type": "object",
"properties": {
"level": {
"type": "string",
"enum": ["DEBUG", "INFO", "WARNING", "ERROR"],
"default": "INFO",
"description": "Logging level."
}
},
"additionalProperties": false
},
"template_engine": {
"type": "string",
"enum": ["JINJA2", "NONE"],
"default": "JINJA2",
"description": "Template engine for string interpolation."
},
"unsafe": {
"type": "boolean",
"default": false,
"description": "When true, allows actors to perform operations flagged as unsafe. Requires --unsafe CLI flag."
},
"default_actor": {
"type": "string",
"description": "Name of the default actor when multiple actors are defined."
}
},
"additionalProperties": false
},
"actors": {
"$ref": "#/$defs/actorMap"
},
"agents": {
"$ref": "#/$defs/actorMap",
"description": "Alias for 'actors'. Both keys are accepted; use one or the other."
},
"routes": {
"type": "object",
"description": "Map of route names to their definitions. Routes connect actors via stream or graph topologies.",
"additionalProperties": {
"$ref": "#/$defs/route"
}
},
"merges": {
"type": "array",
"description": "Stream merge operations combining multiple streams into one.",
"items": {
"type": "object",
"properties": {
"sources": {
"type": "array",
"items": { "type": "string" },
"description": "Source stream names to merge."
},
"target": {
"type": "string",
"description": "Target stream name for merged output."
}
},
"required": ["sources", "target"],
"additionalProperties": false
}
},
"splits": {
"type": "array",
"description": "Stream split operations dividing one stream into multiple.",
"items": {
"type": "object",
"properties": {
"source": {
"type": "string",
"description": "Source stream name to split."
},
"targets": {
"type": "array",
"items": { "type": "string" },
"description": "Target stream names for split output."
}
},
"required": ["source", "targets"],
"additionalProperties": false
}
},
"templates": {
"type": "object",
"description": "Reusable template definitions for Jinja2 template inheritance.",
"additionalProperties": true
},
"instances": {
"type": "object",
"description": "Instantiated templates with bound parameters.",
"additionalProperties": true
},
"global_context": {
"type": "object",
"description": "Key-value pairs available to all actors via {{ context.key }} in templates.",
"additionalProperties": true
},
"prompts": {
"type": "object",
"description": "Named prompt templates that can be referenced by actors.",
"additionalProperties": { "type": "string" }
},
"pipelines": {
"type": "object",
"description": "Hybrid pipeline definitions combining stream and graph stages.",
"additionalProperties": {
"type": "object",
"properties": {
"stages": {
"type": "array",
"items": {
"type": "object",
"properties": {
"name": { "type": "string" },
"type": { "type": "string" },
"config": { "type": "object", "additionalProperties": true }
},
"required": ["name", "type"]
}
},
"metadata": { "type": "object", "additionalProperties": true }
},
"required": ["stages"]
}
}
},
"required": ["name"],
"oneOf": [
{ "required": ["actors"] },
{ "required": ["agents"] }
],
"additionalProperties": false,
"$defs": {
"actorMap": {
"type": "object",
"description": "Map of actor names to their definitions.",
"additionalProperties": {
"$ref": "#/$defs/actorDefinition"
}
},
"actorDefinition": {
"type": "object",
"description": "A single actor definition.",
"properties": {
"type": {
"type": "string",
"enum": ["llm", "tool"],
"description": "Actor type: 'llm' for language model actors, 'tool' for tool-based actors."
},
"config": {
"type": "object",
"description": "Actor configuration. Fields depend on actor type.",
"properties": {
"provider": {
"type": "string",
"description": "LLM provider identifier: openai, anthropic, google, azure, openrouter, etc."
},
"model": {
"type": "string",
"description": "Model identifier within the provider: gpt-4, claude-3.5-sonnet, gemini-pro, etc."
},
"actor": {
"type": "string",
"description": "Combined provider/model format (e.g., 'anthropic/claude-3.5-sonnet'). Alternative to specifying provider and model separately."
},
"system_prompt": {
"type": "string",
"description": "System prompt text. Supports Jinja2 template syntax for dynamic content."
},
"temperature": {
"type": "number",
"minimum": 0.0,
"maximum": 2.0,
"description": "Sampling temperature. Lower values are more deterministic."
},
"max_tokens": {
"type": "integer",
"minimum": 1,
"description": "Maximum number of tokens in the generated response."
},
"memory_enabled": {
"type": "boolean",
"default": false,
"description": "Enable conversation memory for multi-turn interactions."
},
"max_history": {
"type": "integer",
"default": 50,
"minimum": 1,
"description": "Maximum number of conversation turns retained in memory."
},
"unsafe": {
"type": "boolean",
"default": false,
"description": "Allow this specific actor to perform unsafe operations."
},
"options": {
"type": "object",
"description": "Provider-specific options passed through to the underlying LLM API.",
"additionalProperties": true
},
"tools": {
"type": "array",
"description": "List of inline tool definitions for tool-type actors.",
"items": {
"type": "object",
"properties": {
"name": {
"type": "string",
"description": "Tool name."
},
"code": {
"type": "string",
"description": "Inline Python code defining the tool's behavior."
}
},
"required": ["name", "code"],
"additionalProperties": false
}
}
},
"additionalProperties": false
}
},
"required": ["type", "config"],
"additionalProperties": false
},
"route": {
"type": "object",
"description": "A route definition — either a stream or a graph topology.",
"properties": {
"type": {
"type": "string",
"enum": ["stream", "graph"],
"description": "Route type."
},
"stream_type": {
"type": "string",
"enum": ["cold", "hot", "replay"],
"default": "cold",
"description": "Stream type (stream routes only)."
},
"operators": {
"type": "array",
"description": "Processing operators (stream routes).",
"items": {
"type": "object",
"properties": {
"type": {
"type": "string",
"enum": ["map", "graph_execute"],
"description": "Operator type."
},
"params": {
"type": "object",
"properties": {
"agent": { "type": "string", "description": "Actor name for map operators." },
"graph": { "type": "string", "description": "Graph route name for graph_execute operators." }
},
"additionalProperties": true
}
},
"required": ["type"]
}
},
"subscriptions": {
"type": "array",
"items": { "type": "string" },
"description": "Input stream subscriptions."
},
"publications": {
"type": "array",
"items": { "type": "string" },
"description": "Output stream publications."
},
"agents": {
"type": "array",
"items": { "type": "string" },
"description": "Actor names used by this route."
},
"initial_value": {
"description": "Initial stream value."
},
"buffer_size": {
"type": "integer",
"default": 10,
"minimum": 1,
"description": "Stream buffer size."
},
"template_config": {
"type": "object",
"description": "Template-specific configuration.",
"additionalProperties": true
},
"bridge": {
"type": "object",
"description": "Bridge configuration for stream-to-graph upgrades.",
"properties": {
"upgrade_conditions": { "type": "object", "additionalProperties": true },
"downgrade_conditions": { "type": "object", "additionalProperties": true },
"state_extractor": { "type": "string" },
"state_flattener": { "type": "string" },
"preserve_subscriptions": { "type": "boolean", "default": true },
"preserve_checkpointing": { "type": "boolean", "default": true }
},
"additionalProperties": false
},
"metadata": {
"type": "object",
"additionalProperties": true
},
"nodes": {
"type": "object",
"description": "Graph nodes (required for graph routes).",
"additionalProperties": {
"$ref": "#/$defs/graphNode"
}
},
"edges": {
"type": "array",
"description": "Graph edges (required for graph routes).",
"items": {
"$ref": "#/$defs/graphEdge"
}
},
"entry_point": {
"type": "string",
"description": "Entry point node name (required for graph routes)."
},
"checkpointing": {
"type": "boolean",
"default": false,
"description": "Enable graph checkpointing."
},
"checkpoint_dir": {
"type": "string",
"description": "Directory for checkpoint storage."
},
"enable_time_travel": {
"type": "boolean",
"default": false,
"description": "Enable time travel debugging."
},
"parallel_execution": {
"type": "boolean",
"default": false,
"description": "Allow parallel node execution."
},
"state_class": {
"type": "string",
"description": "Custom state class name."
}
},
"required": ["type"],
"additionalProperties": false
},
"graphNode": {
"type": "object",
"description": "A graph node definition.",
"properties": {
"type": {
"type": "string",
"enum": ["agent", "function", "tool", "conditional", "subgraph", "start", "end", "message_router"],
"description": "Node type."
},
"agent": { "type": "string", "description": "Actor name for agent nodes." },
"function": { "type": "string", "description": "Function name for function nodes." },
"tools": {
"type": "array",
"items": { "type": "string" },
"description": "Tool references for tool nodes."
},
"condition": { "type": "object", "additionalProperties": true, "description": "Condition for conditional nodes." },
"subgraph": { "type": "string", "description": "Route name for subgraph nodes." },
"retry_policy": { "type": "object", "additionalProperties": true, "description": "Retry configuration." },
"timeout": { "type": "integer", "minimum": 1, "description": "Timeout in seconds." },
"parallel": { "type": "boolean", "default": false, "description": "Allow parallel execution." },
"metadata": { "type": "object", "additionalProperties": true }
},
"required": ["type"],
"additionalProperties": false
},
"graphEdge": {
"type": "object",
"description": "A graph edge connecting two nodes.",
"properties": {
"source": { "type": "string", "description": "Source node name." },
"target": { "type": "string", "description": "Target node name." },
"condition": { "type": "object", "additionalProperties": true, "description": "Edge condition for conditional routing." },
"metadata": { "type": "object", "additionalProperties": true }
},
"required": ["source", "target"],
"additionalProperties": false
}
}
}
Informal YAML Schema
The following annotated YAML provides an easier-to-read overview of the same schema defined formally above. It is not a validatable schema — it uses informal annotations (comments, <placeholder> syntax, and | alternatives) to describe the structure. Refer to the JSON Schema above for the authoritative, machine-validatable definition.
# ─── Name ─────────────────────────────────────────────────────────── name: <namespace>/<actor-name> # Namespaced actor name (required). Used as the registered name. # Must follow <namespace>/<name> format (e.g., local/reviewer).# ─── Metadata ─────────────────────────────────────────────────────── cleveragents: version: "3.0" # Schema version (optional, default: latest) logging: level: "INFO" # Log level: DEBUG, INFO, WARNING, ERROR (optional) template_engine: "JINJA2" # Template engine: JINJA2 or NONE (optional, default: JINJA2) unsafe: false # Allow unsafe operations (optional, default: false) default_actor: <actor_name> # Default actor to use when multiple are defined (optional)
# ─── Actor Definitions ────────────────────────────────────────────── # The top-level key can be either "actors" or "agents" (both accepted). actors: <actor_name>: type: llm | tool # Actor type (required) config: # ── For type: llm ────────────────────────────────────────── provider: <string> # Provider identifier: openai, anthropic, google, azure, openrouter, etc. (required for LLM) model: <string> # Model identifier: gpt-4, claude-3.5-sonnet, etc. (required for LLM) # OR use the combined format: actor: "<provider>/<model>" # Combined provider/model (alternative to provider + model)
<span style="color: cyan; font-weight: 600;">system_prompt</span>: |<span style="opacity: 0.7;"> # System prompt text, supports Jinja2 templates (optional)</span> You are a helpful assistant. <span style="color: cyan; font-weight: 600;">Project</span>: {{ context.project_name }} <span style="color: cyan; font-weight: 600;">temperature</span>: <span style="color: yellow;">0.7</span><span style="opacity: 0.7;"> # Sampling temperature 0.0-2.0 (optional, default: provider default)</span> <span style="color: cyan; font-weight: 600;">max_tokens</span>: <span style="color: yellow;">4096</span><span style="opacity: 0.7;"> # Maximum output tokens (optional, default: provider default)</span> <span style="color: cyan; font-weight: 600;">memory_enabled</span>: <span style="color: magenta; font-weight: 600;">true</span><span style="opacity: 0.7;"> # Enable conversation memory (optional, default: false)</span> <span style="color: cyan; font-weight: 600;">max_history</span>: <span style="color: yellow;">50</span><span style="opacity: 0.7;"> # Maximum conversation turns to retain (optional, default: 50)</span> <span style="color: cyan; font-weight: 600;">unsafe</span>: <span style="color: magenta; font-weight: 600;">false</span><span style="opacity: 0.7;"> # Allow unsafe operations for this actor (optional, default: false)</span> <span style="color: cyan; font-weight: 600;">options</span>: # Additional provider-specific options (optional) <span style="color: cyan; font-weight: 600;">top_p</span>: <span style="color: yellow;">1.0</span> <span style="color: cyan; font-weight: 600;">frequency_penalty</span>: <span style="color: yellow;">0.0</span> <span style="color: cyan; font-weight: 600;">presence_penalty</span>: <span style="color: yellow;">0.0</span> <span style="color: cyan; font-weight: 600;">stop_sequences</span>: [<span style="color: #66cc66;">"END"</span>] <span style="color: cyan; font-weight: 600;">seed</span>: <span style="color: yellow;">42</span># ── For type: tool ───────────────────────────────────────── tools: - name: <tool_name> # Tool name (required per tool) code: | # Inline Python code (required per tool) def run(input_data): return {"result": input_data["query"]}
# ─── Routes ───────────────────────────────────────────────────────── routes: <route_name>: type: stream | graph # Route type (required)
# ── Stream route fields ────────────────────────────────────── stream_type: cold | hot | replay # Stream type (optional, default: cold) operators: # Processing operators (optional) - type: map | graph_execute # Operator type params: # Operator parameters agent: <actor_name> # Actor to use (for map) graph: <route_name> # Graph to execute (for graph_execute) subscriptions: # Input subscriptions (optional) - <stream_name> publications: # Output publications (optional) - <stream_name> agents: # Actors used by this route (optional) - <actor_name> initial_value: <any> # Initial stream value (optional) buffer_size: 10 # Stream buffer size (optional, default: 10) template_config: {} # Template-specific configuration (optional) bridge: # Bridge configuration for stream↔graph upgrades (optional) upgrade_conditions: {} # Conditions to upgrade from stream to graph downgrade_conditions: {} # Conditions to downgrade from graph to stream state_extractor: <string> # Function to extract state during upgrade state_flattener: <string> # Function to flatten state during downgrade preserve_subscriptions: true # Keep subscriptions during transition (optional) preserve_checkpointing: true # Keep checkpoints during transition (optional) metadata: {} # Arbitrary metadata (optional)
# ── Graph route fields ─────────────────────────────────────── nodes: # Graph nodes (required for graph routes) <node_name>: type: agent | function | tool | conditional | subgraph | start | end | message_router agent: <actor_name> # Actor for agent nodes function: <function_name> # Function for function nodes tools: # Tools for tool nodes - <tool_ref> condition: {} # Condition for conditional nodes subgraph: <route_name> # Subgraph for subgraph nodes retry_policy: {} # Retry configuration (optional) timeout: 30 # Timeout in seconds (optional) parallel: false # Allow parallel execution (optional) metadata: {} # Arbitrary metadata (optional) edges: # Graph edges (required for graph routes) - source: <node_name> # Source node (required) target: <node_name> # Target node (required) condition: {} # Edge condition (optional) metadata: {} # Arbitrary metadata (optional) entry_point: <node_name> # Entry point node (required for graph routes) checkpointing: false # Enable graph checkpointing (optional, default: false) checkpoint_dir: <path> # Checkpoint directory (optional) enable_time_travel: false # Enable time travel debugging (optional, default: false) parallel_execution: false # Allow parallel node execution (optional, default: false) state_class: <string> # Custom state class name (optional) metadata: {} # Arbitrary metadata (optional)
# ─── Merges & Splits ──────────────────────────────────────────────── merges: # Stream merge definitions (optional)
- sources: # Source streams to merge
- <stream_name> target: <stream_name> # Target stream for merged output
splits: # Stream split definitions (optional)
- source: <stream_name> # Source stream to split targets: # Target streams for split output
- <stream_name>
# ─── Templates & Context ──────────────────────────────────────────── templates: {} # Reusable template definitions (optional) instances: {} # Template instances (optional) global_context: {} # Global context available to all actors (optional) prompts: {} # Named prompt templates (optional)
# ─── Pipelines ────────────────────────────────────────────────────── pipelines: # Hybrid pipeline definitions (optional) <pipeline_name>: stages: - name: <stage_name> type: <stage_type> config: {} metadata: {}
Structure Reference
Top-Level Keys
| Key | Type | Required | Description |
|---|---|---|---|
cleveragents |
object | No | Metadata block containing version, logging, template engine, and safety settings. |
actors (or agents) |
object | Yes | Map of actor names to their definitions. Both actors and agents are accepted as the key name. |
routes |
object | No | Map of route names to their definitions. Routes connect actors via stream or graph topologies. |
merges |
list | No | Stream merge operations combining multiple streams into one. |
splits |
list | No | Stream split operations dividing one stream into multiple. |
templates |
object | No | Reusable template definitions for Jinja2 template inheritance. |
instances |
object | No | Instantiated templates with bound parameters. |
global_context |
object | No | Key-value pairs available to all actors via {{ context.key }} in templates. |
prompts |
object | No | Named prompt templates that can be referenced by actors. |
pipelines |
object | No | Hybrid pipelines combining stream and graph stages into sequential workflows. |
cleveragents Block
| Field | Type | Default | Description |
|---|---|---|---|
version |
string | "3.0" |
Schema version for this configuration file. |
logging.level |
string | "INFO" |
Logging level: DEBUG, INFO, WARNING, ERROR. |
template_engine |
string | "JINJA2" |
Template engine for string interpolation: JINJA2 or NONE. |
unsafe |
boolean | false |
When true, allows actors to perform operations flagged as unsafe. Requires --unsafe CLI flag. |
default_actor |
string | (first actor) | Name of the default actor when multiple actors are defined. |
Actor Definition Fields
| Field | Type | Required | Description |
|---|---|---|---|
type |
string | Yes | Actor type. llm for language model actors, tool for tool-based actors. |
config.provider |
string | Yes (LLM) | LLM provider identifier: openai, anthropic, google, azure, openrouter, etc. |
config.model |
string | Yes (LLM) | Model identifier within the provider: gpt-4, claude-3.5-sonnet, gemini-pro, etc. |
config.actor |
string | No | Combined provider/model format. Alternative to specifying provider and model separately. |
config.system_prompt |
string | No | System prompt text. Supports Jinja2 template syntax for dynamic content. |
config.temperature |
float | Provider default | Sampling temperature (0.0 to 2.0). Lower values are more deterministic. |
config.max_tokens |
integer | Provider default | Maximum number of tokens in the generated response. |
config.memory_enabled |
boolean | false |
Enable conversation memory for multi-turn interactions. |
config.max_history |
integer | 50 |
Maximum number of conversation turns retained in memory. |
config.unsafe |
boolean | false |
Allow this specific actor to perform unsafe operations. |
config.options |
object | {} |
Provider-specific options passed through to the underlying LLM API. |
config.tools |
list | Yes (tool) | List of inline tool definitions for tool-type actors. Each tool has name and code. |
Node Types (Graph Routes)
| Type | Description |
|---|---|
agent |
A node backed by an actor. Requires agent field pointing to an actor name. |
function |
A node backed by a Python function. Requires function field. |
tool |
A node that invokes one or more tools. Requires tools list. |
conditional |
A branching node that routes based on conditions. Requires condition field. |
subgraph |
A node that delegates to another graph route. Requires subgraph field. |
start |
The explicit start node (alternative to entry_point). |
end |
A terminal node that ends graph execution. |
message_router |
Routes messages to different nodes based on message content or metadata. |
Examples
Example 1: Minimal Chat Actor (Simple)
A bare-minimum actor configuration for a simple conversational agent:
# minimal-chat.yaml
# Register: agents actor add --config minimal-chat.yaml
name: local/chat
actors:
chat:
type: llm
config:
provider: anthropic
model: claude-3.5-sonnet
system_prompt: "You are a helpful assistant."
routes:
main:
type: stream
operators:
- type: map
params:
agent: chat
publications:
- output
merges:
sources: [output] target: final
This is the simplest possible actor: one LLM actor, one stream route, one merge. No graph complexity, no tools, no templates.
Example 2: Code Reviewer with Tools (Medium)
An actor that can read files and search code to perform code reviews:
# code-reviewer.yaml
# Register: agents actor add --config code-reviewer.yaml
name: local/reviewer
cleveragents:
version: "3.0"
logging:
level: "INFO"
actors:
reviewer:
type: llm
config:
provider: openai
model: gpt-4
temperature: 0.2
max_tokens: 4096
system_prompt: |
You are a senior code reviewer. Analyze code for:
- Security vulnerabilities
- Performance issues
- Code style violations
- Missing error handling
Provide specific, actionable feedback with file and line references.
file_reader:
type: tool
config:
tools:
- name: read_file
code: |
def run(input_data):
path = input_data.get("path", "")
with open(path, "r") as f:
return {"content": f.read(), "path": path}
routes:
review_graph:
type: graph
nodes:
analyze:
type: agent
agent: reviewer
read:
type: tool
tools: [read_file]
report:
type: agent
agent: reviewer
edges:
- source: analyze
target: read
- source: read
target: report
- source: analyze
target: report
condition:
no_files_needed: true
entry_point: analyze
checkpointing: false
output_stream:
type: stream
operators:
- type: graph_execute
params:
graph: review_graph
publications:
- review_output
merges:
sources: [review_output] target: final
This actor defines an LLM reviewer and a file-reading tool, connected via a graph route with conditional edges.
Example 3: Multi-Agent Research Pipeline with Templates (Complex)
A complex actor configuration with multiple LLM actors, Jinja2 templates, environment variable interpolation, memory, and hybrid routing:
# research-pipeline.yaml # Register: agents actor add --config research-pipeline.yaml --unsafename: local/research
cleveragents: version: "3.0" logging: level: "DEBUG" template_engine: "JINJA2" unsafe: true default_actor: orchestrator
actors: orchestrator: type: llm config: provider: anthropic model: claude-3.5-sonnet temperature: 0.7 max_tokens: 8192 memory_enabled: true max_history: 100 system_prompt: | You are an orchestrator managing a research pipeline for {{ context.project_name }}. Topic: {{ context.research_topic }}
Your job is to: 1. Break down the research question into sub-questions 2. Delegate to specialist researchers 3. Synthesize findings into a coherent report {% if context.deadline %} <span style="color: cyan; font-weight: 600;">Deadline</span>: {{ context.deadline }}. Prioritize breadth over depth. {% endif %}researcher: type: llm config: provider: openai model: gpt-4 temperature: 0.3 max_tokens: 4096 system_prompt: | You are a domain expert researcher. Provide thorough, factual analysis. Always cite sources and note confidence levels.
synthesizer: type: llm config: provider: anthropic model: claude-3.5-sonnet temperature: 0.5 max_tokens: 16384 system_prompt: | You are an expert at synthesizing multiple research reports into coherent narratives. Resolve contradictions, highlight consensus, and note gaps.
web_searcher: type: tool config: tools: - name: search_web code: | import os, json def run(input_data): api_key = os.environ.get("SEARCH_API_KEY", "") query = input_data.get("query", "") # Simulated web search return {"results": [], "query": query}
- <span style="color: cyan; font-weight: 600;">name</span>: fetch_url <span style="color: cyan; font-weight: 600;">code</span>: | def run(input_data): url = input_data.get("url", "") return {"content": f"Content from {url}", "url": url}routes: research_graph: type: graph nodes: plan: type: agent agent: orchestrator research_parallel: type: agent agent: researcher parallel: true search: type: tool tools: [search_web, fetch_url] synthesize: type: agent agent: synthesizer review: type: agent agent: orchestrator edges: - source: plan target: research_parallel - source: plan target: search - source: research_parallel target: synthesize - source: search target: synthesize - source: synthesize target: review - source: review target: plan condition: needs_more_research: true entry_point: plan checkpointing: true checkpoint_dir: "${CHECKPOINT_DIR:/tmp/research_checkpoints}" enable_time_travel: true parallel_execution: true
progress_stream: type: stream stream_type: hot operators: - type: map params: agent: orchestrator subscriptions: - research_updates publications: - progress_output buffer_size: 50
output_stream: type: stream operators: - type: graph_execute params: graph: research_graph publications: - research_output
merges:
- sources: [progress_output, research_output] target: final
global_context: project_name: "AI Safety Research" research_topic: "Alignment techniques in large language models" deadline: "2026-03-01"
prompts: deep_dive: "Provide an in-depth analysis of {topic} with at least 5 sources." summary: "Summarize the following research in 500 words: {content}"
templates: research_section: template: | ## {{ section_title }} {{ section_content }} Confidence: {{ confidence_level }} Sources: {{ sources | join(', ') }}
This configuration demonstrates multi-actor orchestration, graph routing with parallel execution, hot stream monitoring, Jinja2 templates with conditionals, environment variables, checkpointing, and global context.
Example 4: Simple Echo/Tool Actor (Simple)
A non-LLM actor that processes input with pure Python code:
# echo-actor.yaml
# Register: agents actor add --config echo-actor.yaml
name: local/echo
cleveragents:
version: "3.0"
actors:
echo:
type: tool
config:
tools:
- name: echo_tool
code: |
def run(input_data):
message = input_data.get("content", "")
return {"response": f"Echo: {message}"}
routes:
main:
type: stream
operators:
- type: map
params:
agent: echo
publications:
- output
merges:
sources: [output] target: final
Example 5: Graph Actor with Conditional Routing (Medium)
An actor that routes between different specialists based on input classification:
# classifier-router.yaml
# Register: agents actor add --config classifier-router.yaml
name: local/smart-router
actors:
classifier:
type: llm
config:
provider: openai
model: gpt-4
temperature: 0.0
max_tokens: 100
system_prompt: |
Classify the user's request into exactly one category:
CODING, WRITING, ANALYSIS, or GENERAL.
Respond with only the category name.
coding_expert:
type: llm
config:
provider: anthropic
model: claude-3.5-sonnet
temperature: 0.2
system_prompt: "You are an expert software engineer. Write clean, well-tested code."
writing_expert:
type: llm
config:
provider: openai
model: gpt-4
temperature: 0.7
system_prompt: "You are a professional writer. Produce clear, engaging prose."
analyst:
type: llm
config:
provider: anthropic
model: claude-3.5-sonnet
temperature: 0.3
system_prompt: "You are a data analyst. Provide thorough, evidence-based analysis."
generalist:
type: llm
config:
provider: openai
model: gpt-4
temperature: 0.5
system_prompt: "You are a helpful general-purpose assistant."
routes:
router_graph:
type: graph
nodes:
classify:
type: agent
agent: classifier
code:
type: agent
agent: coding_expert
write:
type: agent
agent: writing_expert
analyze:
type: agent
agent: analyst
general:
type: agent
agent: generalist
edges:
- source: classify
target: code
condition:
category: "CODING"
- source: classify
target: write
condition:
category: "WRITING"
- source: classify
target: analyze
condition:
category: "ANALYSIS"
- source: classify
target: general
condition:
category: "GENERAL"
entry_point: classify
main:
type: stream
operators:
- type: graph_execute
params:
graph: router_graph
publications:
- output
merges:
sources: [output] target: final
Skill Configuration Files
!!! adr "Architecture Decision" The skill configuration schema, tool references, and skill inclusion are defined in ADR-012: Skill System.
Skill configuration files define reusable, namespaced collections of tools. Skills assemble tools by referencing named tools from the Tool Registry, defining anonymous inline tools, and including other skills. Skills are registered via agents skill add --config <file>.
JSON Schema
The following is the formal JSON Schema definition for skill configuration files.
{
"$schema": "https://json-schema.org/draft/2020-12/schema",
"$id": "https://cleveragents.dev/schemas/skill-config.json",
"title": "CleverAgents Skill Configuration",
"description": "Configuration file schema for defining CleverAgents skills — reusable, namespaced collections of tools.",
"type": "object",
"properties": {
"name": {
"type": "string",
"pattern": "^[a-zA-Z0-9_-]+/[a-zA-Z0-9_-]+$",
"description": "Fully qualified skill name in <namespace>/<name> format."
},
"description": {
"type": "string",
"description": "Human-readable description of what this skill provides."
},
"tools": {
"type": "array",
"description": "References to named tools from the Tool Registry.",
"items": {
"type": "object",
"properties": {
"name": {
"type": "string",
"pattern": "^[a-zA-Z0-9_-]+/[a-zA-Z0-9_-]+$",
"description": "Fully qualified name of a tool registered in the Tool Registry."
},
"description": {
"type": "string",
"description": "Override the tool's registered description within this skill context."
},
"writes": {
"type": "boolean",
"description": "Override the tool's writes capability flag."
},
"checkpointable": {
"type": "boolean",
"description": "Override the tool's checkpointable capability flag."
}
},
"required": ["name"],
"additionalProperties": false
}
},
"inline_tools": {
"type": "array",
"description": "Anonymous tool definitions that exist only within this skill. Not registered in the Tool Registry.",
"items": {
"type": "object",
"properties": {
"name": {
"type": "string",
"description": "Tool name, unique within this skill."
},
"description": {
"type": "string",
"description": "Human-readable description of the tool's purpose."
},
"source": {
"type": "string",
"const": "custom",
"description": "Source type. Always 'custom' for inline tools."
},
"code": {
"type": "string",
"description": "Python code defining the tool's behavior. Must contain a run(input_data) function."
},
"input_schema": {
"$ref": "https://json-schema.org/draft/2020-12/schema",
"description": "JSON Schema describing the tool's input parameters."
},
"writes": {
"type": "boolean",
"default": false,
"description": "Whether this tool performs write operations."
},
"checkpointable": {
"type": "boolean",
"default": false,
"description": "Whether this tool supports checkpointing."
},
"side_effects": {
"type": "array",
"items": { "type": "string" },
"default": [],
"description": "Descriptions of side effects (e.g., 'network_call', 'schema_mutation')."
}
},
"required": ["name", "source", "code"],
"additionalProperties": false
}
},
"includes": {
"type": "array",
"description": "Other skills whose tools are merged into this skill.",
"items": {
"type": "object",
"properties": {
"name": {
"type": "string",
"pattern": "^[a-zA-Z0-9_-]+/[a-zA-Z0-9_-]+$",
"description": "Fully qualified skill name to include."
},
"description": {
"type": "string",
"description": "Override the included skill's description."
}
},
"required": ["name"],
"additionalProperties": false
}
},
"mcp_servers": {
"type": "array",
"description": "MCP server specifications for exposing remote tools.",
"items": {
"$ref": "#/$defs/mcpServer"
}
},
"agent_skill_folders": {
"type": "array",
"description": "Agent Skills Standard folders to include.",
"items": {
"type": "object",
"properties": {
"path": {
"type": "string",
"description": "Path to the folder containing SKILL.md."
},
"name": {
"type": "string",
"description": "Override the skill bundle name."
}
},
"required": ["path"],
"additionalProperties": false
}
}
},
"required": ["name"],
"additionalProperties": false,
"$defs": {
"mcpServer": {
"type": "object",
"description": "An MCP server specification.",
"properties": {
"name": {
"type": "string",
"description": "Identifier for the MCP server."
},
"transport": {
"type": "string",
"enum": ["stdio", "sse", "streamable-http"],
"description": "Transport protocol."
},
"command": {
"type": "string",
"description": "Command to start the server (required for stdio transport)."
},
"args": {
"type": "array",
"items": { "type": "string" },
"description": "Command-line arguments for the server command."
},
"env": {
"type": "object",
"additionalProperties": { "type": "string" },
"description": "Environment variables to set when starting the server."
},
"url": {
"type": "string",
"format": "uri",
"description": "Server URL (required for sse and streamable-http transports)."
},
"headers": {
"type": "object",
"additionalProperties": { "type": "string" },
"description": "HTTP headers for remote server connections."
},
"tool_filter": {
"type": "object",
"properties": {
"include": {
"type": "array",
"items": { "type": "string" },
"description": "Whitelist of tool names to expose."
},
"exclude": {
"type": "array",
"items": { "type": "string" },
"description": "Blacklist of tool names to hide."
}
},
"additionalProperties": false
}
},
"required": ["name", "transport"],
"additionalProperties": false
}
}
}
Informal YAML Schema
The following annotated YAML provides an easier-to-read overview of the same schema defined formally above. It is not a validatable schema — it uses informal annotations to describe the structure. Refer to the JSON Schema above for the authoritative, machine-validatable definition.
# ─── Skill Metadata ─────────────────────────────────────────────────
name: <namespace>/<name> # Fully qualified skill name (required)
description: <string> # Human-readable description (optional)
# ─── Tool References ────────────────────────────────────────────────
# Named tools from the Tool Registry, referenced by fully-qualified name.
tools:
- name: <namespace>/<tool_name> # Reference to a registered tool (required)
description: <string> # Override the tool's description (optional)
writes: <boolean> # Override the tool's writes flag (optional)
checkpointable: <boolean> # Override the tool's checkpointable flag (optional)
# ─── Inline (Anonymous) Tools ───────────────────────────────────────
# Tools defined directly within this skill. These are NOT registered
# in the Tool Registry and cannot be reused outside this skill.
inline_tools:
- name: <string> # Tool name (unique within this skill, required)
description: <string> # Tool description (optional)
source: custom # Source type (required, always "custom" for inline)
code: | # Inline Python code (required)
def run(input_data):
return {"result": "value"}
input_schema: # JSON Schema for tool inputs (optional)
type: object
properties:
param_name:
type: string
description: "Parameter description"
required: ["param_name"]
writes: false # Whether this tool writes (optional, default: false)
checkpointable: false # Whether this tool supports checkpointing (optional, default: false)
side_effects: [] # List of side effect descriptions (optional)
# ─── Included Skills ────────────────────────────────────────────────
# Other skills whose tools are merged into this skill.
includes:
- name: <namespace>/<skill_name> # Fully qualified skill name to include (required)
description: <string> # Override the included skill's description (optional)
# ─── MCP Server Specifications ──────────────────────────────────────
# MCP servers whose tools are exposed through this skill.
mcp_servers:
- name: <string> # Server name for identification (required)
transport: stdio | sse | streamable-http # Transport protocol (required)
command: <string> # Server command (required for stdio)
args: # Command arguments (optional)
- <string>
env: # Environment variables for the server (optional)
KEY: "value"
url: <string> # Server URL (required for sse/streamable-http)
headers: {} # HTTP headers (optional, for sse/streamable-http)
tool_filter: # Filter which tools to expose (optional)
include: # Include only these tools (optional)
- <tool_name>
exclude: # Exclude these tools (optional)
- <tool_name>
# ─── Agent Skills Standard Folders ──────────────────────────────────
# References to Agent Skills Standard (SKILL.md-based) tool bundles.
agent_skill_folders:
path: <string> # Path to the folder containing SKILL.md (required) name: <string> # Override the skill bundle name (optional)
Structure Reference
Top-Level Fields
| Field | Type | Required | Description |
|---|---|---|---|
name |
string | Yes | Fully qualified skill name in <namespace>/<name> format. This value is used as the skill's registered name. |
description |
string | No | Human-readable description of what this skill provides. |
tools |
list | No | References to named tools from the Tool Registry. |
inline_tools |
list | No | Anonymous tool definitions that exist only within this skill. |
includes |
list | No | Other skills to include (their tools are merged in). |
mcp_servers |
list | No | MCP server specifications for exposing remote tools. |
agent_skill_folders |
list | No | Agent Skills Standard folders to include. |
Tool Reference Fields (tools[])
| Field | Type | Required | Description |
|---|---|---|---|
name |
string | Yes | Fully qualified name of a tool registered in the Tool Registry. |
description |
string | No | Override the tool's registered description within this skill context. |
writes |
boolean | No | Override the tool's writes capability flag. |
checkpointable |
boolean | No | Override the tool's checkpointable capability flag. |
Inline Tool Fields (inline_tools[])
| Field | Type | Required | Description |
|---|---|---|---|
name |
string | Yes | Tool name, unique within this skill. |
description |
string | No | Human-readable description of the tool's purpose. |
source |
string | Yes | Always custom for inline tools. |
code |
string | Yes | Python code defining the tool's behavior. Must contain a run(input_data) function. |
input_schema |
object | No | JSON Schema describing the tool's input parameters. |
writes |
boolean | No | Whether this tool performs write operations. Default: false. |
checkpointable |
boolean | No | Whether this tool supports checkpointing. Default: false. |
side_effects |
list | No | Descriptions of any side effects (e.g., ["network_call"], ["schema_mutation"]). |
MCP Server Fields (mcp_servers[])
| Field | Type | Required | Description |
|---|---|---|---|
name |
string | Yes | Identifier for the MCP server. |
transport |
string | Yes | Transport protocol: stdio, sse, or streamable-http. |
command |
string | Conditional | Command to start the server (required for stdio transport). |
args |
list | No | Command-line arguments for the server command. |
env |
object | No | Environment variables to set when starting the server. |
url |
string | Conditional | Server URL (required for sse and streamable-http transports). |
headers |
object | No | HTTP headers for remote server connections. |
tool_filter.include |
list | No | Whitelist of tool names to expose. When set, only these tools are available. |
tool_filter.exclude |
list | No | Blacklist of tool names to hide. Applied after include. |
Examples
Example 1: Minimal Read-Only Skill (Simple)
A skill that bundles a few built-in file operations:
# file-reader-skill.yaml
# Register: agents skill add --config file-reader-skill.yaml
name: local/file-reader
description: "Basic file reading operations"
tools:
- name: builtin/read_file
- name: builtin/list_directory
name: builtin/search_files
Example 2: Git Operations Skill with MCP (Medium)
A skill that combines built-in git tools with GitHub tools via an MCP server:
# git-and-github-skill.yaml
# Register: agents skill add --config git-and-github-skill.yaml
name: local/git-github
description: "Git operations and GitHub integration"
tools:
- name: builtin/git_status
- name: builtin/git_diff
- name: builtin/git_log
- name: builtin/git_blame
includes:
- name: local/file-reader
mcp_servers:
name: github transport: stdio command: npx args:- "-y"
"@modelcontextprotocol/server-github" env: GITHUB_PERSONAL_ACCESS_TOKEN: "${GITHUB_TOKEN}" tool_filter: include:- create_issue
- create_pull_request
- list_repos
get_file_contents
This skill includes another skill (local/file-reader) and adds GitHub tools from an MCP server, filtered to only expose four specific tools.
Example 3: Full DevOps Toolkit (Complex)
A comprehensive skill with multiple includes, MCP servers, inline tools, and agent skill folders:
# devops-toolkit.yaml
# Register: agents skill add --config devops-toolkit.yaml
name: local/devops-toolkit
description: "Full-stack development and operations toolkit"
tools:
- name: builtin/shell_execute
description: "Execute shell commands in the project sandbox"
- name: local/validate-api-compat
description: "Check API backward compatibility"
includes:
- name: local/file-reader
- name: local/git-github
- name: local/docker-tools
inline_tools:
-
name: run_migrations
description: "Run database migrations with rollback support"
source: custom
code: |
import subprocess
def run(input_data):
direction = input_data.get("direction", "up")
count = input_data.get("count", 1)
result = subprocess.run(
["alembic", direction, str(count)],
capture_output=True, text=True
)
return {
"success": result.returncode == 0,
"stdout": result.stdout,
"stderr": result.stderr
}
input_schema:
type: object
properties:
direction:
type: string
enum: ["up", "down"]
description: "Migration direction"
count:
type: integer
default: 1
description: "Number of migrations to run"
required: ["direction"]
writes: true
checkpointable: true
side_effects: ["schema_mutation"]
-
name: health_check
description: "Check service health endpoints"
source: custom
code: |
import urllib.request
def run(input_data):
url = input_data.get("url", "http://localhost:8000/health")
try:
resp = urllib.request.urlopen(url, timeout=10)
return {"status": resp.status, "healthy": resp.status == 200}
except Exception as e:
return {"status": 0, "healthy": False, "error": str(e)}
input_schema:
type: object
properties:
url:
type: string
description: "Health check URL"
writes: false
mcp_servers:
- name: linear
transport: stdio
command: npx
args: ["-y", "@modelcontextprotocol/server-linear"]
env:
LINEAR_API_KEY: "${LINEAR_API_KEY}"
agent_skill_folders:
- path: ./skills/deploy-to-staging name: deploy-staging
path: ./skills/code-review-bundle name: code-review
Example 4: Skill with Only Inline Tools (Simple)
A self-contained skill that defines everything inline, with no external dependencies:
# text-processing-skill.yaml
# Register: agents skill add --config text-processing-skill.yaml
name: local/text-processing
description: "Simple text transformation utilities"
inline_tools:
-
name: word_count description: "Count words in text" source: custom code: | def run(input_data): text = input_data.get("text", "") return {"count": len(text.split())} input_schema: type: object properties: text: type: string description: "Text to count words in" required: ["text"] writes: false
-
name: to_uppercase description: "Convert text to uppercase" source: custom code: | def run(input_data): return {"result": input_data.get("text", "").upper()} input_schema: type: object properties: text: type: string required: ["text"] writes: false
name: extract_urls description: "Extract URLs from text" source: custom code: | import re def run(input_data): text = input_data.get("text", "") urls = re.findall(r'https?://[^\s<>"{}|\^`[]]+', text) return {"urls": urls, "count": len(urls)} writes: false
Action Configuration Files
!!! adr "Architecture Decision" The action configuration schema and action-to-plan lifecycle are defined in ADR-006: Plan Lifecycle.
Action configuration files define reusable plan templates — complete specifications for work that can be applied to projects. Actions are created via agents action create --config <file>.
JSON Schema
The following is the formal JSON Schema definition for action configuration files.
{
"$schema": "https://json-schema.org/draft/2020-12/schema",
"$id": "https://cleveragents.dev/schemas/action-config.json",
"title": "CleverAgents Action Configuration",
"description": "Configuration file schema for defining CleverAgents actions — reusable plan templates that specify work to be applied to projects.",
"type": "object",
"properties": {
"name": {
"type": "string",
"pattern": "^[a-zA-Z0-9_-]+/[a-zA-Z0-9_-]+$",
"description": "Fully qualified action name in <namespace>/<name> format."
},
"description": {
"type": "string",
"description": "Short (one-line) description of the action."
},
"long_description": {
"type": "string",
"description": "Detailed multi-line description explaining purpose, usage, and expected outcomes."
},
"strategy_actor": {
"type": "string",
"pattern": "^[a-zA-Z0-9_-]+/[a-zA-Z0-9_-]+$",
"description": "Actor to use during the Strategize phase. Must reference a registered actor."
},
"execution_actor": {
"type": "string",
"pattern": "^[a-zA-Z0-9_-]+/[a-zA-Z0-9_-]+$",
"description": "Actor to use during the Execute phase. Must reference a registered actor."
},
"estimation_actor": {
"type": "string",
"pattern": "^[a-zA-Z0-9_-]+/[a-zA-Z0-9_-]+$",
"description": "Actor for effort and cost estimation before execution."
},
"review_actor": {
"type": "string",
"pattern": "^[a-zA-Z0-9_-]+/[a-zA-Z0-9_-]+$",
"description": "Actor for reviewing execution results."
},
"apply_actor": {
"type": "string",
"pattern": "^[a-zA-Z0-9_-]+/[a-zA-Z0-9_-]+$",
"description": "Actor to use during the Apply phase."
},
"invariant_actor": {
"type": "string",
"pattern": "^[a-zA-Z0-9_-]+/[a-zA-Z0-9_-]+$",
"description": "Actor for reconciling conflicting invariants across scopes."
},
"definition_of_done": {
"type": "string",
"description": "Clear, measurable criteria that define when the action's work is complete."
},
"reusable": {
"type": "boolean",
"default": true,
"description": "Whether the action persists after being used."
},
"read_only": {
"type": "boolean",
"default": false,
"description": "Whether the action is restricted to read-only operations."
},
"state": {
"type": "string",
"enum": ["available", "archived"],
"default": "available",
"description": "State of the action."
},
"arguments": {
"type": "array",
"description": "Typed parameters that users supply when using the action via agents plan use --arg name=value.",
"items": {
"$ref": "#/$defs/argument"
}
},
"automation_profile": {
"type": "string",
"description": "Default automation profile for plans created from this action."
},
"invariants": {
"type": "array",
"items": { "type": "string" },
"description": "Constraints carried forward as plan-level invariants when the action is used."
}
},
"required": ["name", "description", "strategy_actor", "execution_actor", "definition_of_done"],
"additionalProperties": false,
"$defs": {
"argument": {
"type": "object",
"description": "A typed parameter for the action.",
"properties": {
"name": {
"type": "string",
"description": "Argument name. Used as the key in --arg name=value."
},
"type": {
"type": "string",
"enum": ["string", "integer", "float", "boolean", "list"],
"description": "Data type of the argument."
},
"required": {
"type": "boolean",
"default": false,
"description": "Whether the argument must be provided."
},
"description": {
"type": "string",
"description": "Human-readable description shown in help text."
},
"default": {
"description": "Default value when the argument is not provided. Type must match the 'type' field."
},
"validation_pattern": {
"type": "string",
"description": "Regex pattern for validating string arguments."
},
"min_value": {
"type": "number",
"description": "Minimum acceptable value for integer and float arguments."
},
"max_value": {
"type": "number",
"description": "Maximum acceptable value for integer and float arguments."
}
},
"required": ["name", "type"],
"additionalProperties": false
}
}
}
Informal YAML Schema
The following annotated YAML provides an easier-to-read overview of the same schema defined formally above. It is not a validatable schema — it uses informal annotations to describe the structure. Refer to the JSON Schema above for the authoritative, machine-validatable definition.
# ─── Action Identity ────────────────────────────────────────────────
name: <namespace>/<name> # Fully qualified action name (required)
description: <string> # Short description (required)
long_description: | # Detailed description (optional)
Multi-line detailed explanation of what this action does,
when to use it, and what outcomes to expect.
# ─── Lifecycle Actors ───────────────────────────────────────────────
strategy_actor: <namespace>/<name> # Actor for the Strategize phase (required)
execution_actor: <namespace>/<name> # Actor for the Execute phase (required)
estimation_actor: <namespace>/<name> # Actor for effort/cost estimation (optional)
review_actor: <namespace>/<name> # Actor for reviewing results (optional)
apply_actor: <namespace>/<name> # Actor for the Apply phase (optional)
invariant_actor: <namespace>/<name> # Invariant Reconciliation Actor (optional)
# ─── Completion Criteria ────────────────────────────────────────────
definition_of_done: | # Criteria for when the action is complete (required)
Clear, measurable criteria that define success.
Multiple criteria can be listed.
# ─── Action Properties ──────────────────────────────────────────────
reusable: true # Keep action after use (optional, default: true)
read_only: false # Restrict to read-only operations (optional, default: false)
state: available # State of the action: available or archived (optional, default: available)
# ─── Arguments ──────────────────────────────────────────────────────
# Arguments are typed parameters that must be supplied when using
# the action via agents plan use --arg name=value.
arguments:
- name: <string> # Argument name (required)
type: string | integer | float | boolean | list # Argument type (required)
required: true # Whether the argument must be provided (optional, default: false)
description: <string> # Human-readable description (optional)
default: <value> # Default value when not provided (optional)
validation_pattern: <regex> # Regex pattern for string validation (optional)
min_value: <number> # Minimum value for numeric types (optional)
max_value: <number> # Maximum value for numeric types (optional)
# ─── Automation ─────────────────────────────────────────────────────
automation_profile: <string> # Default automation profile name (optional)
# ─── Invariants ─────────────────────────────────────────────────────
# Invariants carried forward as plan-level invariants when this action is used.
invariants:
<string> # Invariant text
Structure Reference
Top-Level Fields
| Field | Type | Required | Description |
|---|---|---|---|
name |
string | Yes | Fully qualified action name in <namespace>/<name> format. This value is used as the action's registered name. |
description |
string | Yes | Short (one-line) description of the action. |
long_description |
string | No | Detailed multi-line description explaining purpose, usage, and expected outcomes. |
strategy_actor |
string | Yes | Actor to use during the Strategize phase. Must reference a registered actor. |
execution_actor |
string | Yes | Actor to use during the Execute phase. Must reference a registered actor. |
estimation_actor |
string | No | Actor to use for effort and cost estimation before execution. |
review_actor |
string | No | Actor to use for reviewing execution results. |
apply_actor |
string | No | Actor to use during the Apply phase. |
invariant_actor |
string | No | Actor for reconciling conflicting invariants across scopes. Carried forward to plans. |
definition_of_done |
string | Yes | Clear, measurable criteria that define when the action's work is complete. |
reusable |
boolean | No | Whether the action persists after being used. Default: true. |
read_only |
boolean | No | Whether the action is restricted to read-only operations. Default: false. |
state |
string | No | State of the action: available or archived. Default: available. |
arguments |
list | No | Typed parameters that users supply when using the action. |
automation_profile |
string | No | Default automation profile for plans created from this action. |
invariants |
list | No | Constraints carried forward as plan-level invariants when the action is used. |
Argument Fields (arguments[])
| Field | Type | Required | Description |
|---|---|---|---|
name |
string | Yes | Argument name. Used as the key in --arg name=value. |
type |
string | Yes | Data type: string, integer, float, boolean, list. |
required |
boolean | No | Whether the argument must be provided. Default: false. |
description |
string | No | Human-readable description shown in help text and agents action show. |
default |
any | No | Default value when the argument is not provided. Type must match type. |
validation_pattern |
string | No | Regex pattern for validating string arguments. |
min_value |
number | No | Minimum acceptable value for integer and float arguments. |
max_value |
number | No | Maximum acceptable value for integer and float arguments. |
Examples
Example 1: Simple Linting Action (Simple)
A minimal action for running linters on a project:
# lint-check.yaml # Create: agents action create --config lint-check.yamlname: local/lint-check description: "Run linting checks on the project"
strategy_actor: local/strategist execution_actor: local/executor
definition_of_done: | All linting checks pass with zero errors.
reusable: true read_only: true state: available
Usage: agents plan use local/lint-check local/api-service
Example 2: Code Coverage Action with Arguments (Medium)
An action with typed arguments, invariants, and a specific automation profile:
# code-coverage.yaml
# Create: agents action create --config code-coverage.yaml
name: local/code-coverage
description: "Increase test coverage to a target percentage"
long_description: |
Analyzes the current test coverage of a project, identifies
modules with low coverage, and generates comprehensive test
suites to meet the target coverage percentage.
The action prioritizes:
- Business-critical modules (auth, payments)
- Recently modified code
- Error-prone areas based on git history
strategy_actor: local/strategist
execution_actor: local/executor
estimation_actor: local/estimator
definition_of_done: |
Test coverage reaches the target_coverage_percent threshold
across all specified modules. All generated tests pass.
No existing tests are broken by the changes.
reusable: true
read_only: false
arguments:
-
name: target_coverage_percent
type: integer
required: true
description: "Target test coverage percentage (1-100)"
min_value: 1
max_value: 100
-
name: test_command
type: string
required: false
description: "Test framework command to use"
default: "pytest --cov"
-
name: exclude_patterns
type: list
required: false
description: "File patterns to exclude from coverage analysis"
default: ["/migrations/", "**/conftest.py"]
-
name: focus_modules
type: list
required: false
description: "Specific modules to prioritize for coverage"
automation_profile: trusted
invariants:
- "Generated tests must not import production secrets or credentials"
- "Test files must follow the project's existing test naming conventions"
"All database interactions in tests must use mocks or fixtures"
Usage: agents plan use local/code-coverage local/api-service --arg target_coverage_percent=85
Example 3: Security Audit Action (Complex)
A comprehensive security audit action with multiple actors, extensive arguments, and strong invariants:
# security-audit.yaml
# Create: agents action create --config security-audit.yaml
name: local/security-audit
description: "Comprehensive security audit of a project"
long_description: |
Performs a thorough security audit covering:
- Dependency vulnerability scanning (CVEs)
- Static Application Security Testing (SAST)
- Authentication and authorization review
- Input validation and injection prevention
- Secrets detection and credential scanning
- API security (rate limiting, CORS, headers)
- Data handling and privacy compliance
Generates a detailed report with severity ratings (Critical,
High, Medium, Low, Informational) and remediation guidance.
Optionally creates fix plans for identified issues.
strategy_actor: local/security-strategist
execution_actor: local/security-scanner
estimation_actor: local/estimator
review_actor: local/security-reviewer
invariant_actor: local/invariant-resolver
definition_of_done: |
All critical and high severity findings have been identified.
A complete security report has been generated with:
- Severity classification for each finding
- Remediation steps for each finding
- Risk score for the overall project
If auto_fix is enabled, all critical findings have remediation
plans created as child plans.
reusable: true
read_only: false
state: available
arguments:
-
name: severity_threshold
type: string
required: false
description: "Minimum severity to include in report"
default: "low"
validation_pattern: "^(critical|high|medium|low|informational)$"
-
name: auto_fix
type: boolean
required: false
description: "Automatically create fix plans for critical findings"
default: false
-
name: scan_dependencies
type: boolean
required: false
description: "Include dependency vulnerability scanning"
default: true
-
name: compliance_frameworks
type: list
required: false
description: "Compliance frameworks to check against"
default: ["owasp-top-10"]
-
name: max_findings
type: integer
required: false
description: "Maximum number of findings to report"
default: 100
min_value: 1
max_value: 1000
-
name: ignore_paths
type: list
required: false
description: "Paths to exclude from scanning"
default: ["/node_modules/", "/vendor/", "/.git/"]
automation_profile: supervised
invariants:
- "Never modify production database schemas during audit"
- "Never execute discovered exploit code against live systems"
- "All findings must include reproducible steps"
- "Secrets found during scanning must be redacted in reports"
"Remediation fixes must not break existing tests"
Usage: agents plan use local/security-audit local/api-service local/web-app --arg auto_fix=true --arg severity_threshold=medium
Example 4: Database Migration Action (Medium)
An action for managing database schema migrations:
# db-migrate.yaml
# Create: agents action create --config db-migrate.yaml
name: local/db-migrate
description: "Plan and execute database schema migrations"
strategy_actor: local/db-strategist
execution_actor: local/db-executor
definition_of_done: |
All migration scripts are generated and pass dry-run validation.
Rollback scripts are generated for each migration.
The migration can be applied without data loss.
reusable: true
read_only: false
arguments:
-
name: migration_tool
type: string
required: false
description: "Migration tool to use"
default: "alembic"
validation_pattern: "^(alembic|flyway|django|knex)$"
-
name: dry_run
type: boolean
required: false
description: "Only generate and validate, do not apply"
default: true
-
name: target_schema
type: string
required: false
description: "Target schema version identifier"
automation_profile: supervised
invariants:
- "All migrations must include corresponding rollback scripts"
- "Data migrations must preserve existing data integrity"
"Schema changes must maintain backward compatibility for 24 hours"
Tool Configuration Files
!!! adr "Architecture Decision" The tool configuration schema, source types, and capability metadata are defined in ADR-011: Tool System.
Tool configuration files define namespaced, independently registered, callable operations. Tools are the atomic unit of execution — each has a name, input/output schemas, capability metadata, and an implementation. Tools are registered via agents tool add --config <file> and can then be referenced by name in skills and actor graphs.
JSON Schema
The following is the formal JSON Schema definition for tool configuration files.
{
"$schema": "https://json-schema.org/draft/2020-12/schema",
"$id": "https://cleveragents.dev/schemas/tool-config.json",
"title": "CleverAgents Tool Configuration",
"description": "Configuration file schema for defining CleverAgents tools — namespaced, independently registered, callable operations.",
"type": "object",
"properties": {
"name": {
"type": "string",
"pattern": "^[a-zA-Z0-9_-]+/[a-zA-Z0-9_-]+$",
"description": "Fully qualified tool name in <namespace>/<name> format."
},
"description": {
"type": "string",
"description": "Human-readable description of the tool's purpose and behavior."
},
"source": {
"type": "string",
"enum": ["custom", "mcp", "agent_skill", "builtin"],
"description": "Tool source type. Determines which implementation fields are required."
},
"code": {
"type": "string",
"description": "Python code implementing the tool. Required when source is 'custom'. Must define a run(input_data) function."
},
"mcp_server": {
"type": "string",
"description": "Name of the MCP server exposing this tool. Required when source is 'mcp'."
},
"mcp_tool_name": {
"type": "string",
"description": "Name of the tool on the MCP server. Required when source is 'mcp'."
},
"agent_skill_path": {
"type": "string",
"description": "Path to the Agent Skills Standard folder containing SKILL.md. Required when source is 'agent_skill'."
},
"input_schema": {
"$ref": "https://json-schema.org/draft/2020-12/schema",
"description": "JSON Schema describing the tool's input parameters."
},
"output_schema": {
"$ref": "https://json-schema.org/draft/2020-12/schema",
"description": "JSON Schema describing the tool's output format."
},
"writes": {
"type": "boolean",
"default": false,
"description": "Whether the tool performs any write operations."
},
"write_scope": {
"type": "string",
"description": "Scope of write operations (e.g., 'filesystem', 'database:migrations', 'api:github')."
},
"checkpointable": {
"type": "boolean",
"default": false,
"description": "Whether the tool supports checkpointing — saving state before execution and restoring on failure."
},
"checkpoint_scope": {
"type": "string",
"enum": ["file", "transaction", "snapshot", "composite"],
"description": "The checkpointing strategy."
},
"side_effects": {
"type": "array",
"items": { "type": "string" },
"default": [],
"description": "Side effects that cannot be undone by checkpointing alone (e.g., 'network_call', 'schema_mutation', 'email_sent')."
},
"idempotent": {
"type": "boolean",
"default": false,
"description": "Whether the tool is safe to retry — calling it multiple times with the same input produces the same result."
},
"read_only": {
"type": "boolean",
"default": false,
"description": "Whether the tool only reads data without modifying anything."
},
"unsafe": {
"type": "boolean",
"default": false,
"description": "Whether the tool is flagged as unsafe. Requires allow_unsafe_tools: true in the automation profile."
},
"timeout": {
"type": "integer",
"default": 300,
"minimum": 1,
"description": "Default execution timeout in seconds."
},
"resource_slots": {
"type": "array",
"description": "Declared resource dependencies for this tool.",
"items": {
"$ref": "#/$defs/resourceSlot"
}
},
"lifecycle": {
"type": "object",
"description": "Lifecycle hook implementations.",
"properties": {
"discover": {
"type": "string",
"description": "Python code or function name run during tool discovery."
},
"activate": {
"type": "string",
"description": "Python code or function name run when the tool is activated."
},
"deactivate": {
"type": "string",
"description": "Python code or function name run when the tool is deactivated."
}
},
"additionalProperties": false
},
},
"required": ["name", "description", "source"],
"allOf": [
{
"if": { "properties": { "source": { "const": "custom" } } },
"then": { "required": ["code"] }
},
{
"if": { "properties": { "source": { "const": "mcp" } } },
"then": { "required": ["mcp_server", "mcp_tool_name"] }
},
{
"if": { "properties": { "source": { "const": "agent_skill" } } },
"then": { "required": ["agent_skill_path"] }
}
],
"additionalProperties": false,
"$defs": {
"resourceSlot": {
"type": "object",
"description": "A declared resource dependency slot.",
"properties": {
"name": {
"type": "string",
"description": "Identifier for this resource slot."
},
"resource_type": {
"type": "string",
"description": "The resource type this slot requires (e.g., 'git-checkout', 'fs-directory')."
},
"access": {
"type": "string",
"enum": ["read_only", "read_write"],
"description": "Access level needed."
},
"description": {
"type": "string",
"description": "Human-readable description of how the tool uses this resource."
},
"binding": {
"type": "string",
"enum": ["contextual", "static", "parameter"],
"default": "contextual",
"description": "How the slot is resolved at runtime."
},
"static_resource": {
"type": "string",
"pattern": "^[a-zA-Z0-9_-]+/[a-zA-Z0-9_-]+$",
"description": "Fully qualified name of a specific resource. Required when binding is 'static'."
}
},
"required": ["name", "resource_type", "access"],
"if": { "properties": { "binding": { "const": "static" } } },
"then": { "required": ["name", "resource_type", "access", "static_resource"] },
"additionalProperties": false
}
}
}
Informal YAML Schema
The following annotated YAML provides an easier-to-read overview of the same schema defined formally above. It is not a validatable schema — it uses informal annotations to describe the structure. Refer to the JSON Schema above for the authoritative, machine-validatable definition.
# ─── Tool Identity ────────────────────────────────────────────────── name: <namespace>/<name> # Fully qualified tool name (required) description: <string> # Human-readable description (required)# ─── Source and Implementation ────────────────────────────────────── source: custom | mcp | agent_skill | builtin # Tool source type (required)
# For source: custom — inline Python implementation code: | # Python code with a run(input_data) function (required for custom) def run(input_data): param = input_data.get("param_name", "default") # ... tool logic ... return {"result": "value"}
# For source: mcp — tool exposed by an MCP server mcp_server: <string> # MCP server name (required for mcp) mcp_tool_name: <string> # Tool name on the MCP server (required for mcp)
# For source: agent_skill — tool from an Agent Skills Standard folder agent_skill_path: <string> # Path to the SKILL.md folder (required for agent_skill)
# ─── Input/Output Schema ──────────────────────────────────────────── input_schema: # JSON Schema for tool inputs (optional but recommended) type: object properties: param_name: type: string description: "Parameter description" enum: ["value1", "value2"] # Enumerated valid values (optional) default: "value1" # Default value (optional) numeric_param: type: integer description: "A numeric parameter" minimum: 0 # Minimum value (optional) maximum: 100 # Maximum value (optional) required: ["param_name"] # Required parameters
output_schema: # JSON Schema for tool outputs (optional) type: object properties: result: type: string success: type: boolean
# ─── Capability Metadata ──────────────────────────────────────────── writes: false # Whether the tool performs write operations (optional, default: false) write_scope: <string> # Scope of writes, e.g. "filesystem", "database:migrations" (optional) checkpointable: false # Whether the tool supports checkpointing (optional, default: false) checkpoint_scope: <string> # Checkpointing strategy: "file", "transaction", "snapshot", "composite" (optional) side_effects: # List of side effect types (optional)
- <string> # e.g. "network_call", "schema_mutation", "process_spawn" idempotent: false # Whether the tool is safe to retry (optional, default: false) read_only: false # Whether the tool only reads (optional, default: false) unsafe: false # Whether the tool is flagged as unsafe (optional, default: false) timeout: 300 # Default execution timeout in seconds (optional, default: 300)
# ─── Resource Bindings ────────────────────────────────────────────── # Declare which resources this tool operates on. resource_slots:
- name: <string> # Slot name for reference (required) resource_type: <string> # Required resource type, e.g. "git-checkout", "fs-directory" (required) access: read_only | read_write # Access level needed (required) description: <string> # Description of how the resource is used (optional) binding: contextual | static | parameter # How the slot is resolved (optional, default: contextual) static_resource: <namespace>/<name> # Specific resource (required when binding is static)
# ─── Lifecycle Hooks ──────────────────────────────────────────────── lifecycle: discover: <string> # Python function or code for tool discovery (optional) activate: <string> # Python function or code run when tool is activated (optional) deactivate: <string> # Python function or code run when tool is deactivated (optional)
Structure Reference
Top-Level Fields
| Field | Type | Required | Description |
|---|---|---|---|
name |
string | Yes | Fully qualified tool name in <namespace>/<name> format. This value is used as the tool's registered name. |
description |
string | Yes | Human-readable description of the tool's purpose and behavior. |
source |
string | Yes | Tool source type. Determines which implementation fields are required: custom (inline Python), mcp (MCP server), agent_skill (SKILL.md folder), builtin (first-party operations). |
code |
string | Conditional | Python code implementing the tool. Required when source is custom. Must define a run(input_data) function that accepts a dict and returns a dict. |
mcp_server |
string | Conditional | Name of the MCP server exposing this tool. Required when source is mcp. |
mcp_tool_name |
string | Conditional | Name of the tool on the MCP server. Required when source is mcp. |
agent_skill_path |
string | Conditional | Path to the Agent Skills Standard folder containing SKILL.md. Required when source is agent_skill. |
input_schema |
object | No | JSON Schema describing the tool's input parameters. Strongly recommended for all tools. |
output_schema |
object | No | JSON Schema describing the tool's output format. |
| Capability Metadata Fields |
| Field | Type | Default | Description |
|---|---|---|---|
writes |
boolean | false |
Whether the tool performs any write operations (file writes, database mutations, API calls with side effects). |
write_scope |
string | — | Describes the scope of write operations, e.g. "filesystem", "database:migrations", "api:github". Used for access checks and audit logging. |
checkpointable |
boolean | false |
Whether the tool supports checkpointing — the ability to save state before execution and restore it on failure. |
checkpoint_scope |
string | — | The checkpointing strategy: file (file-level backup), transaction (database transaction), snapshot (full state snapshot), composite (multi-resource checkpoint). |
side_effects |
list | [] |
Descriptions of side effects that cannot be undone by checkpointing alone, e.g. "network_call", "schema_mutation", "email_sent", "process_spawn". |
idempotent |
boolean | false |
Whether the tool is safe to retry — calling it multiple times with the same input produces the same result. |
read_only |
boolean | false |
Whether the tool only reads data without modifying anything. When true, writes must be false. |
unsafe |
boolean | false |
Whether the tool is flagged as unsafe. Unsafe tools require allow_unsafe_tools: true in the automation profile. |
timeout |
integer | 300 |
Default execution timeout in seconds. Can be overridden per invocation. |
Resource Slot Fields (resource_slots[])
| Field | Type | Required | Description |
|---|---|---|---|
name |
string | Yes | Identifier for this resource slot, used for reference in code and logs. |
resource_type |
string | Yes | The resource type this slot requires, e.g. git-checkout, fs-directory, fs-file. |
access |
string | Yes | Access level: read_only or read_write. |
description |
string | No | Human-readable description of how the tool uses this resource. |
binding |
string | No | How the slot is resolved at runtime. contextual (default): resolved from the plan's project. static: hardcoded to a specific resource. parameter: passed as a tool argument at invocation time. |
static_resource |
string | Conditional | Fully qualified name of a specific resource. Required when binding is static. |
Lifecycle Hook Fields (lifecycle)
| Field | Type | Required | Description |
|---|---|---|---|
discover |
string | No | Code or function name run during tool discovery to check prerequisites and report capabilities. |
activate |
string | No | Code or function name run when the tool is activated (e.g., establish connections, validate credentials). |
deactivate |
string | No | Code or function name run when the tool is deactivated (e.g., close connections, release resources). |
Examples
Example 1: Simple Read-Only Tool (Simple)
A tool that reads a file and returns its contents:
# read-config-file.yaml # Register: agents tool add --config read-config-file.yamlname: local/read-config description: "Read and parse a configuration file (JSON, YAML, or TOML)"
source: custom code: | import json, os def run(input_data): path = input_data["path"] if not os.path.exists(path): return {"error": f"File not found: {path}", "success": False} with open(path, "r") as f: content = f.read() ext = os.path.splitext(path)[1].lower() if ext == ".json": parsed = json.loads(content) elif ext in (".yaml", ".yml"): import yaml parsed = yaml.safe_load(content) else: parsed = None return {"content": content, "parsed": parsed, "path": path, "success": True}
input_schema: type: object properties: path: type: string description: "Path to the configuration file" required: ["path"]
output_schema: type: object properties: content: type: string parsed: type: object success: type: boolean
writes: false read_only: true idempotent: true
Example 2: Database Migration Tool (Medium)
A tool that runs database migrations with rollback support:
# run-migrations.yaml
# Register: agents tool add --config run-migrations.yaml
name: local/run-migrations
description: "Run database migrations with direction control and rollback support"
source: custom
code: |
import subprocess
def run(input_data):
direction = input_data.get("direction", "up")
count = input_data.get("count", 1)
dry_run = input_data.get("dry_run", False)
cmd = ["alembic"]
if dry_run:
cmd.append("<span style="color: cyan;">--sql</span>")
cmd.extend([direction, str(count)])
result = subprocess.run(cmd, capture_output=True, text=True, timeout=120)
return {
<span style="color: cyan; font-weight: 600;">"success"</span>: result.returncode == 0,
<span style="color: cyan; font-weight: 600;">"stdout"</span>: result.stdout,
<span style="color: cyan; font-weight: 600;">"stderr"</span>: result.stderr,
<span style="color: cyan; font-weight: 600;">"direction"</span>: direction,
<span style="color: cyan; font-weight: 600;">"count"</span>: count,
<span style="color: cyan; font-weight: 600;">"dry_run"</span>: dry_run,
<span style="color: cyan; font-weight: 600;">"return_code"</span>: result.returncode
}
input_schema:
type: object
properties:
direction:
type: string
enum: ["up", "down"]
description: "Migration direction: 'up' to apply, 'down' to rollback"
count:
type: integer
default: 1
minimum: 1
maximum: 50
description: "Number of migrations to run"
dry_run:
type: boolean
default: false
description: "Generate SQL without executing"
required: ["direction"]
writes: true
write_scope: "database:migrations"
checkpointable: true
checkpoint_scope: "transaction"
side_effects:
- "schema_mutation"
idempotent: false
timeout: 120
resource_slots:
name: database resource_type: "local/database" access: read_write description: "The database to run migrations against" binding: contextual
Example 3: MCP-Backed GitHub Tool (Simple)
A tool that delegates to a GitHub MCP server:
# github-create-issue.yaml
# Register: agents tool add --config github-create-issue.yaml
name: local/github-create-issue
description: "Create a GitHub issue via the GitHub MCP server"
source: mcp
mcp_server: github
mcp_tool_name: create_issue
input_schema:
type: object
properties:
owner:
type: string
description: "Repository owner"
repo:
type: string
description: "Repository name"
title:
type: string
description: "Issue title"
body:
type: string
description: "Issue body (Markdown)"
labels:
type: array
items:
type: string
description: "Labels to apply"
required: ["owner", "repo", "title"]
writes: true
write_scope: "api:github"
checkpointable: false
side_effects:
"network_call" idempotent: false
Example 4: Deployment Tool with Resource Bindings (Complex)
A comprehensive deployment tool with multiple resource slots, lifecycle hooks, and safety metadata:
# deploy-staging.yaml # Register: agents tool add --config deploy-staging.yamlname: local/deploy-staging description: "Deploy the current build to the staging environment with health checks"
source: custom code: | import subprocess, time, urllib.request, json
def run(input_data): service = input_data["service"] version = input_data.get("version", "latest") wait_healthy = input_data.get("wait_healthy", True) health_timeout = input_data.get("health_timeout", 120)
# Build the deployment command cmd = ["kubectl", "set", "image", f"deployment/{service}", f"{service}={service}:{version}", "--namespace=staging"] result = subprocess.run(cmd, capture_output=True, text=True, timeout=60) if result.returncode != 0: return {"success": False, "error": result.stderr, "phase": "deploy"}
# Wait for rollout rollout = subprocess.run( ["kubectl", "rollout", "status", f"deployment/{service}", "--namespace=staging", f"--timeout={health_timeout}s"], capture_output=True, text=True, timeout=health_timeout + 10 ) if rollout.returncode != 0: return {"success": False, "error": rollout.stderr, "phase": "rollout"}
# Health check if wait_healthy: health_url = f"http://{service}.staging.svc.cluster.local/health" start = time.time() while time.time() - start < health_timeout: try: resp = urllib.request.urlopen(health_url, timeout=5) if resp.status == 200: body = json.loads(resp.read()) if body.get("status") == "healthy": return { "success": True, "service": service, "version": version, "health": body } except Exception: pass time.sleep(5) return {"success": False, "error": "Health check timeout", "phase": "health"}
return {"success": True, "service": service, "version": version}input_schema: type: object properties: service: type: string description: "Name of the service to deploy" version: type: string default: "latest" description: "Docker image version tag" wait_healthy: type: boolean default: true description: "Wait for health check to pass" health_timeout: type: integer default: 120 minimum: 10 maximum: 600 description: "Health check timeout in seconds" required: ["service"]
output_schema: type: object properties: success: type: boolean service: type: string version: type: string error: type: string phase: type: string
writes: true write_scope: "infrastructure:kubernetes" checkpointable: true checkpoint_scope: "composite" side_effects:
- "network_call"
- "process_spawn"
- "infrastructure_mutation" idempotent: false unsafe: true timeout: 300
resource_slots:
- name: repo resource_type: git-checkout access: read_only description: "Source repository for build artifacts" binding: contextual
- name: cluster_config resource_type: fs-file access: read_only description: "Kubernetes cluster configuration (kubeconfig)" binding: static static_resource: local/staging-kubeconfig
lifecycle: activate: | def activate(context): import subprocess result = subprocess.run(["kubectl", "cluster-info"], capture_output=True, text=True) if result.returncode != 0: raise RuntimeError("Cannot connect to Kubernetes cluster") return {"cluster": "connected"} deactivate: | def deactivate(context): pass # No cleanup needed
Example 5: API Compatibility Checker (Medium)
A tool that validates API backward compatibility:
# validate-api-compat.yaml
# Register: agents tool add --config validate-api-compat.yaml
name: local/validate-api-compat
description: "Validate that API changes maintain backward compatibility with existing clients"
source: custom
code: |
import subprocess, json
def run(input_data):
spec_path = input_data.get("spec_path", "openapi.yaml")
base_branch = input_data.get("base_branch", "main")
# Get the base spec from the main branch
base_result = subprocess.run(
["git", "show", f"{base_branch}:{spec_path}"],
capture_output=True, text=True
)
if base_result.returncode != 0:
return {"compatible": True, "reason": "No base spec found (new API)", "changes": []}
# Run OpenAPI diff
diff_result = subprocess.run(
["oasdiff", "breaking", "--format", "json",
"--base", "/dev/stdin", "--revision", spec_path],
input=base_result.stdout,
capture_output=True, text=True
)
if diff_result.returncode == 0:
return {"compatible": True, "changes": [], "reason": "No breaking changes"}
<span style="color: cyan; font-weight: 600;">try</span>:
breaking = json.loads(diff_result.stdout)
except json.JSONDecodeError:
breaking = [{"description": diff_result.stdout}]
return {
<span style="color: cyan; font-weight: 600;">"compatible"</span>: False,
<span style="color: cyan; font-weight: 600;">"changes"</span>: breaking,
<span style="color: cyan; font-weight: 600;">"reason"</span>: f"Found {len(breaking)} breaking change(s)"
}
input_schema:
type: object
properties:
spec_path:
type: string
default: "openapi.yaml"
description: "Path to the OpenAPI specification file"
base_branch:
type: string
default: "main"
description: "Branch to compare against for backward compatibility"
required: []
writes: false
read_only: true
checkpointable: false
idempotent: true
timeout: 60
resource_slots:
name: repo resource_type: git-checkout access: read_only description: "Git repository containing the API spec" binding: contextual
Validation Configuration Files
!!! adr "Architecture Decision" The validation configuration schema and read-only enforcement are defined in ADR-013: Validation Abstraction.
Validation configuration files define Validations — a specialized subtype of Tool. A Validation YAML file uses the same format as a Tool YAML file, with an additional validation block containing the validation-specific metadata. Because Validation extends Tool, all tool fields (name, description, source, code, input_schema, output_schema, capability metadata, resource bindings, lifecycle hooks) are available and behave identically. Validations are registered via agents validation add --config <file>.
Read-only enforcement: Validations are always read-only. The writes and checkpointable fields are always forced to false regardless of what the YAML file specifies — if a validation YAML includes writes: true or checkpointable: true, these values are silently overridden to false at registration time. Because this is enforced, it is idiomatic to omit read_only, writes, and checkpointable from validation YAML files entirely (they are unnecessary). The examples below include read_only: true for documentation clarity, but it is not required.
Shared namespace: The name field in a validation YAML occupies the same namespace as tools. Registration will fail if a tool with the same name already exists.
Validation-Specific Fields
In addition to all fields from the Tool Configuration schema, a Validation YAML adds:
| Field | Type | Default | Description |
|---|---|---|---|
validation.mode |
string | "required" |
"required" or "informational". Required validations must pass for execution to proceed; informational validations report results without blocking. |
wraps |
string | (none) | Name of an existing registered Tool to wrap. When set, the Validation delegates execution to the wrapped Tool and passes its output through the transform function. Mutually exclusive with source and code. |
transform |
string | (none) | Python function that converts the wrapped Tool's output to the Validation return format. Required when wraps is set (unless the wrapped Tool already returns validation-format output). The function receives the Tool's output as its argument and must return { "passed": bool, ... }. |
argument_mapping |
object | (none) | Dictionary mapping the wrapped Tool's input parameter names to either a Validation input parameter name (string) or a fixed literal value. When present, only mapped arguments are forwarded to the wrapped Tool. When absent, the Validation's input arguments are passed through unchanged. Only valid when wraps is set. |
When wraps is used, the source and code fields must be omitted — the Validation's implementation is the wrapped Tool plus the transform. The Validation inherits the wrapped Tool's input_schema, resource_slots, and timeout unless explicitly overridden. When the Validation defines a custom input_schema, an argument_mapping should be provided to specify how the Validation's inputs map to the wrapped Tool's expected inputs. See the Tool Wrapping subsection under Core Concepts > Validation for full semantics.
The output of a validation tool must conform to a structured JSON format:
| Field | Type | Required | Description |
|---|---|---|---|
passed |
boolean | Yes | Whether the validation passed (true) or failed (false). |
message |
string | No | Human-readable summary of the validation result. |
data |
object | No | Arbitrary structured data in any format the validation chooses to convey informational output about the result. |
Examples
Example 1: Unit Test Validation (Required)
# validations/run-tests.yaml
# Register: agents validation add --config validations/run-tests.yaml
name: local/run-tests
description: "Run unit tests with coverage and report pass/fail"
source: custom
code: |
import subprocess, json
def run(input_data):
threshold = input_data.get("coverage_threshold", 80)
result = subprocess.run(
["pytest", "--cov=src", f"--cov-fail-under={threshold}", "--tb=short", "-q"],
capture_output=True, text=True
)
passed = result.returncode == 0
return {
"passed": passed,
"message": "All tests passed" if passed else f"Tests failed (exit code {result.returncode})",
"data": {
"stdout": result.stdout,
"stderr": result.stderr,
"returncode": result.returncode,
"coverage_threshold": threshold
}
}
validation:
mode: required
input_schema:
type: object
properties:
coverage_threshold:
type: integer
default: 80
description: "Minimum coverage percentage required"
read_only: true
idempotent: true
timeout: 600
resource_slots:
name: repo resource_type: git-checkout access: read_only binding: contextual
Example 2: Lint Check Validation (Required, Simple)
# validations/lint-check.yamlname: local/lint-check description: "Run linter and report pass/fail"
source: custom code: | import subprocess def run(input_data): result = subprocess.run(["ruff", "check", "."], capture_output=True, text=True) passed = result.returncode == 0 return { "passed": passed, "message": "Lint clean" if passed else f"Lint errors found", "data": {"stdout": result.stdout, "stderr": result.stderr} }
validation: mode: required
read_only: true idempotent: true timeout: 300
Example 3: Bundle Size Check (Informational)
# validations/check-bundle-size.yamlname: local/check-bundle-size description: "Check bundle size (advisory — does not block execution)"
source: custom code: | import subprocess, json def run(input_data): result = subprocess.run(["node", "scripts/check-bundle-size.js"], capture_output=True, text=True) try: size_data = json.loads(result.stdout) except json.JSONDecodeError: size_data = {"raw_output": result.stdout} passed = result.returncode == 0 return { "passed": passed, "message": "Bundle size within limits" if passed else "Bundle size exceeds advisory threshold", "data": size_data }
validation: mode: informational
read_only: true timeout: 120
Example 4: MCP-Backed Validation
A validation that uses an MCP server to perform security scanning:
# validations/security-scan.yamlname: local/security-scan description: "Run security vulnerability scan via MCP"
source: mcp mcp_server: "npx @security/mcp-scanner" mcp_tool_name: scan_vulnerabilities
validation: mode: required
read_only: true timeout: 300
Example 5: Wrapping an Existing Tool
A validation that wraps an existing local/run-tests tool, reusing its implementation and interpreting its output as pass/fail. No test-running logic is duplicated — only the pass/fail interpretation is defined.
# validations/tests-pass.yaml # Assumes local/run-tests is already registered as a Tool that runs # the test suite and returns {returncode, tests_run, tests_passed, ...}name: local/tests-pass description: "Validate that all unit tests pass (wraps local/run-tests)"
wraps: local/run-tests
transform: | def transform(tool_output): passed = tool_output.get("returncode") == 0 tests_run = tool_output.get("tests_run", 0) tests_passed = tool_output.get("tests_passed", 0) return { "passed": passed, "message": f"{tests_passed}/{tests_run} tests passed" if passed else f"{tests_run - tests_passed} tests failed", "data": tool_output }
validation: mode: required
timeout: 600
Note: source, code, input_schema, and resource_slots are omitted — they are inherited from the wrapped local/run-tests tool. Only the transform, validation.mode, and optional overrides (like timeout) are needed. If the Validation defines a custom input_schema, an argument_mapping should also be provided to map the Validation's inputs to the wrapped Tool's expected inputs.
Example 6: Wrapping with Different Interpretation (Informational)
A second validation wrapping the same tool but extracting a different signal — coverage threshold — as an advisory check:
# validations/coverage-check.yamlname: local/coverage-check description: "Check test coverage exceeds threshold (advisory)"
wraps: local/run-tests
transform: | def transform(tool_output): coverage = tool_output.get("coverage_percent", 0) threshold = 80 return { "passed": coverage >= threshold, "message": f"Coverage: {coverage:.1f}% (threshold: {threshold}%)", "data": { "coverage_percent": coverage, "threshold": threshold, "above_threshold": coverage >= threshold } }
validation: mode: informational
Resource Type Configuration Files
!!! adr "Architecture Decision" The resource type system, custom type registration, and type hierarchy are defined in ADR-008: Resource System.
Resource type configuration files define custom resource type schemas that extend the built-in resource types (git-checkout, git, fs-mount, fs-directory, etc.). A resource type defines what CLI arguments agents resource add <type> accepts, whether instances are physical or virtual, allowed child/parent types, auto-discovery behavior, sandbox strategy, and handler implementation. Custom types are registered via agents resource type add --config <file>.
JSON Schema
The following is the formal JSON Schema definition for resource type configuration files.
{
"$schema": "https://json-schema.org/draft/2020-12/schema",
"$id": "https://cleveragents.dev/schemas/resource-type-config.json",
"title": "CleverAgents Resource Type Configuration",
"description": "Configuration file schema for defining custom resource types that extend the built-in resource types.",
"type": "object",
"properties": {
"name": {
"type": "string",
"pattern": "^[a-zA-Z0-9_-]+/[a-zA-Z0-9_-]+$",
"description": "Fully qualified resource type name in <namespace>/<name> format."
},
"description": {
"type": "string",
"description": "Human-readable description of what this resource type represents."
},
"physical": {
"type": "boolean",
"description": "true for physical types (concrete manifestations), false for virtual types (abstract identity linking equivalent physical resources)."
},
"user_addable": {
"type": "boolean",
"default": true,
"description": "Whether users can create instances directly via 'agents resource add <type>'."
},
"cli_args": {
"type": "array",
"description": "Arguments accepted by 'agents resource add <type>'.",
"items": {
"$ref": "#/$defs/cliArg"
}
},
"child_types": {
"type": "array",
"description": "Resource types that can be children of this type.",
"items": {
"$ref": "#/$defs/childType"
}
},
"parent_types": {
"type": "array",
"description": "Resource types that can be parents of this type. If omitted, the resource can be top-level.",
"items": {
"type": "object",
"properties": {
"type": {
"type": "string",
"description": "Parent resource type name."
},
"description": {
"type": "string",
"description": "Description of the parent-child relationship from the child's perspective."
}
},
"required": ["type"],
"additionalProperties": false
}
},
"sandbox_strategy": {
"type": "string",
"enum": ["git_worktree", "copy_on_write", "transaction_rollback", "snapshot", "none"],
"description": "Sandbox strategy for instances of this type."
},
"handler": {
"type": "object",
"description": "The Python handler class that implements resource operations.",
"properties": {
"class": {
"type": "string",
"description": "Python class name that implements the resource handler interface."
},
"module": {
"type": "string",
"description": "Python module path where the handler class is defined."
},
"config": {
"type": "object",
"description": "Arbitrary configuration passed to the handler constructor.",
"additionalProperties": true
}
},
"required": ["class", "module"],
"additionalProperties": false
},
"auto_discovery": {
"type": "object",
"description": "Controls automatic child resource discovery when a resource of this type is created.",
"properties": {
"enabled": {
"type": "boolean",
"default": true,
"description": "Whether auto-discovery runs when a resource of this type is created."
},
"scan_depth": {
"type": "integer",
"default": 3,
"minimum": 1,
"description": "Maximum recursion depth for scanning."
},
"include_patterns": {
"type": "array",
"items": { "type": "string" },
"description": "Glob patterns for resources to discover."
},
"exclude_patterns": {
"type": "array",
"items": { "type": "string" },
"description": "Glob patterns for resources to skip during discovery."
}
},
"additionalProperties": false
},
"equivalence": {
"type": "object",
"description": "Equivalence criteria for virtual types. Only applicable when physical is false.",
"properties": {
"criteria": {
"type": "array",
"items": { "type": "string" },
"description": "Fields used to determine equivalence (e.g., 'content_hash', 'filename', 'permissions', 'url')."
},
"description": {
"type": "string",
"description": "Human-readable description of the equivalence rule."
}
},
"required": ["criteria"],
"additionalProperties": false
},
},
"required": ["name", "description", "physical", "sandbox_strategy", "handler"],
"if": {
"properties": { "physical": { "const": false } }
},
"then": {
"required": ["name", "description", "physical", "sandbox_strategy", "handler", "equivalence"]
},
"additionalProperties": false,
"$defs": {
"cliArg": {
"type": "object",
"description": "A CLI argument definition for 'agents resource add <type>'.",
"properties": {
"name": {
"type": "string",
"description": "Argument name. Becomes --<name> on the CLI."
},
"type": {
"type": "string",
"enum": ["string", "path", "integer", "boolean", "url"],
"description": "Argument type. 'path' validates that the path exists; 'url' validates URL format."
},
"required": {
"type": "boolean",
"default": false,
"description": "Whether the argument must be provided."
},
"description": {
"type": "string",
"description": "Description shown in --help."
},
"default": {
"description": "Default value when the argument is not provided."
},
"validation_pattern": {
"type": "string",
"description": "Regex pattern for validating string and url type arguments."
}
},
"required": ["name", "type"],
"additionalProperties": false
},
"childType": {
"type": "object",
"description": "An allowed child resource type relationship.",
"properties": {
"type": {
"type": "string",
"description": "Child resource type name."
},
"auto_discover": {
"type": "boolean",
"default": false,
"description": "Whether children of this type are automatically created when the parent resource is registered."
},
"manual_link": {
"type": "boolean",
"default": true,
"description": "Whether manual 'agents resource link-child' is allowed for this relationship."
},
"description": {
"type": "string",
"description": "Description of the parent-child relationship."
},
"max_count": {
"type": ["integer", "null"],
"minimum": 1,
"description": "Maximum number of children of this type. null or omitted means unlimited."
}
},
"required": ["type"],
"additionalProperties": false
}
}
}
Informal YAML Schema
The following annotated YAML provides an easier-to-read overview of the same schema defined formally above. It is not a validatable schema — it uses informal annotations to describe the structure. Refer to the JSON Schema above for the authoritative, machine-validatable definition.
# ─── Resource Type Identity ───────────────────────────────────────── name: <namespace>/<name> # Fully qualified resource type name (required) description: <string> # Human-readable description (required)# ─── Type Classification ──────────────────────────────────────────── physical: true # Whether instances are physical or virtual (required) # Physical: a specific, concrete manifestation (this file at this path) # Virtual: an abstract identity linking equivalent physical resources user_addable: true # Whether users can create instances directly (optional, default: true)
# ─── CLI Arguments ────────────────────────────────────────────────── # Define the arguments accepted by
agents resource add <type>. cli_args:
- name: <string> # Argument name (becomes --<name> on CLI) (required) type: string | path | integer | boolean | url # Argument type (required) required: true # Whether the argument is required (optional, default: false) description: <string> # Description shown in help text (optional) default: <value> # Default value (optional) validation_pattern: <regex> # Regex validation for string/url types (optional)
# ─── Parent/Child Type Relationships ──────────────────────────────── # Define which resource types can be children of this type. child_types:
- type: <string> # Child resource type name (required) auto_discover: true # Automatically create children when parent is created (optional, default: false) manual_link: true # Allow manual parent-child linking (optional, default: true) description: <string> # Description of the relationship (optional) max_count: <integer> # Maximum number of children of this type (optional, null = unlimited)
# Define which resource types can be parents of this type. parent_types:
- type: <string> # Parent resource type name (required) description: <string> # Description of the relationship (optional) # If parent_types is omitted, the resource can be top-level (no parent required).
# ─── Sandbox Strategy ────────────────────────────────────────────── sandbox_strategy: <string> # Sandbox strategy for instances of this type (required) # Built-in strategies: "git_worktree", "copy_on_write", # "transaction_rollback", "snapshot", "none"
# ─── Handler Implementation ──────────────────────────────────────── handler: class: <string> # Python class implementing the resource handler (required) module: <string> # Python module path (required) config: {} # Handler-specific configuration (optional)
# ─── Auto-Discovery Configuration ────────────────────────────────── auto_discovery: enabled: true # Whether child auto-discovery runs on creation (optional, default: true) scan_depth: <integer> # Max depth for recursive scanning (optional, default: 3) include_patterns: # Glob patterns for resources to auto-discover (optional) - <string> exclude_patterns: # Glob patterns to exclude from auto-discovery (optional) - <string>
# ─── Virtual Type Configuration ───────────────────────────────────── # Only applicable when physical: false (virtual types). equivalence: criteria: # Fields used to determine equivalence (required for virtual) - <string> # e.g. "content_hash", "filename", "permissions", "url" description: <string> # Human-readable description of the equivalence rule (optional)
Structure Reference
Top-Level Fields
| Field | Type | Required | Description |
|---|---|---|---|
name |
string | Yes | Fully qualified resource type name in <namespace>/<name> format. This value is used as the resource type's registered name. |
description |
string | Yes | Human-readable description of what this resource type represents. |
physical |
boolean | Yes | true for physical types (concrete manifestations), false for virtual types (abstract identity linking equivalent physical resources). |
user_addable |
boolean | No | Whether users can create instances directly via agents resource add <type>. Default: true. When false, instances are only created via auto-discovery as children of other resources. |
sandbox_strategy |
string | Yes | The sandbox strategy used when executing within this resource type. |
handler |
object | Yes | The Python handler class that implements resource operations. |
CLI Argument Fields (cli_args[])
| Field | Type | Required | Description |
|---|---|---|---|
name |
string | Yes | Argument name. Becomes --<name> on the CLI. |
type |
string | Yes | Argument type: string, path, integer, boolean, url. The path type validates that the path exists; url type validates URL format. |
required |
boolean | No | Whether the argument must be provided. Default: false. |
description |
string | No | Description shown in --help and agents resource type show. |
default |
any | No | Default value when the argument is not provided. |
validation_pattern |
string | No | Regex pattern for validating string and url type arguments. |
Child Type Fields (child_types[])
| Field | Type | Required | Description |
|---|---|---|---|
type |
string | Yes | Name of the allowed child resource type (e.g., fs-directory, svn-file). |
auto_discover |
boolean | No | Whether children of this type are automatically created when the parent resource is registered. Default: false. |
manual_link |
boolean | No | Whether manual agents resource link-child is allowed for this relationship. Default: true. |
description |
string | No | Description of the parent-child relationship. |
max_count |
integer | No | Maximum number of children of this type. null or omitted means unlimited. |
Parent Type Fields (parent_types[])
| Field | Type | Required | Description |
|---|---|---|---|
type |
string | Yes | Name of the allowed parent resource type. |
description |
string | No | Description of the parent-child relationship from the child's perspective. |
Handler Fields (handler)
| Field | Type | Required | Description |
|---|---|---|---|
class |
string | Yes | Python class name that implements the resource handler interface. |
module |
string | Yes | Python module path where the handler class is defined. |
config |
object | No | Arbitrary configuration passed to the handler constructor. |
Auto-Discovery Fields (auto_discovery)
| Field | Type | Required | Description |
|---|---|---|---|
enabled |
boolean | No | Whether auto-discovery runs when a resource of this type is created. Default: true. |
scan_depth |
integer | No | Maximum recursion depth for scanning. Default: 3. |
include_patterns |
list | No | Glob patterns for resources to discover. |
exclude_patterns |
list | No | Glob patterns for resources to skip during discovery. |
Equivalence Fields (equivalence) — Virtual Types Only
| Field | Type | Required | Description |
|---|---|---|---|
criteria |
list | Yes (virtual) | Fields used to determine whether two physical resources are equivalent. Common criteria: content_hash, filename, permissions, url, commit_hash, tree_hash. |
description |
string | No | Human-readable description of the equivalence rule. |
Built-in Sandbox Strategies
| Strategy | Description |
|---|---|
git_worktree |
Creates a git worktree for isolated execution. Changes are committed to a branch and merged on apply. |
copy_on_write |
Creates a CoW copy of the resource. Changes are applied by replacing the original. |
transaction_rollback |
Wraps operations in a database transaction. Rolled back on failure, committed on apply. |
snapshot |
Takes a full snapshot before execution. Restored on rollback. |
none |
No sandbox. Changes are applied directly. Use only for read-only or idempotent operations. |
Examples
Example 1: Simple SVN Repository Type (Simple)
A custom resource type for Subversion repositories:
# svn-type.yaml # Register: agents resource type add --config svn-type.yamlname: local/svn description: "A Subversion (SVN) repository checkout"
physical: true user_addable: true
cli_args:
- name: url type: url required: true description: "SVN repository URL" validation_pattern: "^svn(\+ssh)?://.|^https?://."
- name: checkout-path type: path required: true description: "Local checkout directory"
- name: revision type: string required: false description: "Specific revision to checkout" default: "HEAD"
child_types:
- type: fs-directory auto_discover: true description: "Working copy directory tree"
- type: local/svn-revision auto_discover: true description: "SVN revision history"
sandbox_strategy: copy_on_write
handler: class: SVNHandler module: cleveragents.resources.handlers.svn config: svn_binary: "svn" trust_server_cert: true
auto_discovery: enabled: true scan_depth: 2 exclude_patterns: - "/.svn/"
Usage: agents resource add local/svn local/legacy-repo --url svn://svn.example.com/trunk --checkout-path /repos/legacy
Example 2: S3 Bucket Resource Type (Medium)
A resource type for Amazon S3 buckets with prefix-based child discovery:
# s3-bucket-type.yaml # Register: agents resource type add --config s3-bucket-type.yamlname: local/s3-bucket description: "An Amazon S3 bucket with prefix-based object organization"
physical: true user_addable: true
cli_args:
- name: bucket type: string required: true description: "S3 bucket name" validation_pattern: "^[a-z0-9][a-z0-9.-]{1,61}[a-z0-9]$"
- name: region type: string required: false description: "AWS region" default: "us-east-1"
- name: prefix type: string required: false description: "Key prefix to scope resource to a 'subdirectory'" default: ""
- name: profile type: string required: false description: "AWS CLI profile name" default: "default"
child_types:
- type: local/s3-prefix auto_discover: true description: "S3 key prefixes (virtual directories)" max_count: 1000
- type: local/s3-object auto_discover: true description: "S3 objects (files)"
sandbox_strategy: copy_on_write
handler: class: S3BucketHandler module: cleveragents.resources.handlers.s3 config: max_object_size: 104857600 # 100 MB default_acl: "private"
auto_discovery: enabled: true scan_depth: 3 include_patterns: - "/*.json" - "/.yaml" - "**/.yml" - "/*.py" - "/.sql" exclude_patterns: - "**/.log" - "/*.tmp" - "/node_modules/**"
Example 3: Database Resource Type (Medium)
A resource type for relational databases:
# database-type.yaml # Register: agents resource type add --config database-type.yamlname: local/database description: "A relational database (PostgreSQL, MySQL, SQLite)"
physical: true user_addable: true
cli_args:
- name: connection-string type: string required: true description: "Database connection string (e.g., postgresql://user:pass@host/db)"
- name: engine type: string required: false description: "Database engine" default: "postgresql" validation_pattern: "^(postgresql|mysql|sqlite|mssql)$"
- name: read-only type: boolean required: false description: "Connect in read-only mode" default: false
- name: schema type: string required: false description: "Database schema to scope to" default: "public"
child_types:
- type: local/db-table auto_discover: true description: "Database tables"
- type: local/db-view auto_discover: true description: "Database views"
- type: local/db-migration auto_discover: false manual_link: true description: "Migration scripts linked to this database"
parent_types:
- type: git-checkout description: "Repository containing the application that owns this database"
sandbox_strategy: transaction_rollback
handler: class: DatabaseHandler module: cleveragents.resources.handlers.database config: connection_pool_size: 5 statement_timeout: 30000 # 30 seconds log_queries: true
auto_discovery: enabled: true scan_depth: 1
Example 4: Virtual File Type (Simple)
A virtual resource type that links equivalent physical files across repositories:
# virtual-config-file-type.yaml # Register: agents resource type add --config virtual-config-file-type.yamlname: local/config-file description: "Virtual type linking equivalent configuration files across repositories"
physical: false user_addable: false
child_types:
- type: fs-file auto_discover: false manual_link: true description: "Physical file instances"
- type: git-tree-entry auto_discover: false manual_link: true description: "Git tree entries representing the same config file"
sandbox_strategy: none
handler: class: VirtualConfigFileHandler module: cleveragents.resources.handlers.virtual_config config: track_content_drift: true
equivalence: criteria: - content_hash - filename description: "Two physical resources represent the same config file when they share the same filename and content hash"
Example 5: Docker Registry Resource Type (Complex)
A resource type for Docker container registries with full auto-discovery and lifecycle hooks:
# docker-registry-type.yaml # Register: agents resource type add --config docker-registry-type.yamlname: local/docker-registry description: "A Docker container registry with image and tag discovery"
physical: true user_addable: true
cli_args:
- name: registry-url type: url required: true description: "Docker registry URL (e.g., registry.example.com, ghcr.io/org)" validation_pattern: "^[a-zA-Z0-9][a-zA-Z0-9.-]+(:[0-9]+)?(/[a-zA-Z0-9._-]+)*$"
- name: username type: string required: false description: "Registry username for authentication"
- name: password-env type: string required: false description: "Environment variable name containing the registry password"
- name: namespace type: string required: false description: "Image namespace or organization filter"
child_types:
- type: local/docker-image auto_discover: true description: "Docker images in the registry" max_count: 500
- type: local/docker-tag auto_discover: true description: "Image tags"
sandbox_strategy: none
handler: class: DockerRegistryHandler module: cleveragents.resources.handlers.docker config: api_version: "v2" page_size: 100 cache_ttl: 300
auto_discovery: enabled: true scan_depth: 2 include_patterns: - "/latest" - "/main" - "/release-*" exclude_patterns: - "/sha256:" - "**/-dirty"
Context View Configuration
!!! adr "Architecture Decision" The context view configuration, resource filtering, and phase-specific context assembly are defined in ADR-014: Context Management (ACMS).
Context views control how project resources are filtered and presented to actors during different plan phases. They are managed via agents project context set and its associated CLI flags, but can also be defined declaratively in YAML for portability and version control.
JSON Schema
The following is the formal JSON Schema definition for context view configuration files.
{
"$schema": "https://json-schema.org/draft/2020-12/schema",
"$id": "https://cleveragents.dev/schemas/context-view-config.json",
"title": "CleverAgents Context View Configuration",
"description": "Configuration file schema for defining context views that control how project resources are filtered and presented to actors during plan phases.",
"type": "object",
"properties": {
"project": {
"type": "string",
"pattern": "^[a-zA-Z0-9_-]+/[a-zA-Z0-9_-]+$",
"description": "Fully qualified project name this view applies to."
},
"view": {
"type": "string",
"enum": ["default", "strategize", "execute", "apply"],
"description": "Phase view. 'default' is the fallback for all phases; phase-specific views override 'default'."
},
"include_resources": {
"type": "array",
"items": { "type": "string" },
"description": "Whitelist of resource names to include. When specified, only these resources provide context."
},
"exclude_resources": {
"type": "array",
"items": { "type": "string" },
"description": "Blacklist of resource names to exclude. Applied after include_resources."
},
"include_paths": {
"type": "array",
"items": { "type": "string" },
"description": "Glob patterns for file paths to include. When specified, only matching files are included."
},
"exclude_paths": {
"type": "array",
"items": { "type": "string" },
"description": "Glob patterns for file paths to exclude. Applied after include_paths."
},
"hot_max_tokens": {
"type": ["integer", "null"],
"minimum": 1,
"description": "Soft cap on the number of tokens in hot (immediate) context. null means no soft cap."
},
"warm_max_decisions": {
"type": "integer",
"minimum": 1,
"description": "Maximum number of decisions retained in warm context."
},
"cold_max_decisions": {
"type": "integer",
"minimum": 1,
"description": "Maximum number of decisions retained in cold context."
},
"query_limit": {
"type": "integer",
"default": 20,
"minimum": 1,
"description": "Maximum number of retrieval results per query against the cold tier."
},
"max_file_size": {
"type": "integer",
"default": 1048576,
"minimum": 1,
"description": "Maximum individual file size in bytes that will be included in context. Default: 1 MB."
},
"max_total_size": {
"type": "integer",
"default": 52428800,
"minimum": 1,
"description": "Maximum aggregate size in bytes across all included files. Default: 50 MB."
},
"summarize": {
"type": "boolean",
"default": false,
"description": "When true, large context segments are automatically summarized rather than excluded."
},
"summary_max_tokens": {
"type": "integer",
"minimum": 1,
"description": "Maximum tokens for each generated summary. Only applies when summarize is true."
},
"strategy": {
"type": "array",
"items": { "type": "string" },
"description": "Ordered list of ACMS context strategies to use for this view. Overrides the global strategy list. Valid built-in values: simple-keyword, semantic-embedding, breadth-depth-navigator, arce, temporal-archaeology, plan-decision-context."
},
"default_breadth": {
"type": "integer",
"minimum": 0,
"default": 2,
"description": "Default number of hops from focus nodes in the UKO graph for context expansion. 0 = focus nodes only."
},
"default_depth": {
"oneOf": [
{ "type": "integer", "minimum": 0 },
{ "type": "string", "pattern": "^[A-Z][A-Z0-9_]*$" }
],
"default": 3,
"description": "Default detail depth for context fragments. May be a non-negative integer or a named level string from the active domain's DetailLevelMap (e.g., 'SIGNATURES', 'FULL_SOURCE'). Named levels are resolved to integers via the DetailLevelMap inheritance chain. Default: 3."
},
"depth_gradient": {
"type": "object",
"additionalProperties": {
"oneOf": [
{ "type": "integer", "minimum": 0 },
{ "type": "string", "pattern": "^[A-Z][A-Z0-9_]*$" }
]
},
"description": "Per-hop detail depth overrides. Key is the hop distance (0 = focus node), value is an integer depth or named level string. Hops not listed use default_depth."
},
"skeleton_ratio": {
"type": "number",
"minimum": 0.0,
"maximum": 1.0,
"default": 0.15,
"description": "Fraction of the context budget reserved for inherited plan skeleton context. 0.0 = no skeleton inheritance; 1.0 = all budget to skeleton."
},
"temporal_scope": {
"type": "string",
"enum": ["current", "recent", "all"],
"default": "current",
"description": "Temporal scope for UKO node resolution. 'current' = only isCurrent nodes; 'recent' = current + nodes valid within warm retention window; 'all' = include historical versions."
},
"auto_refresh": {
"type": "boolean",
"default": true,
"description": "When true, the ACMS automatically re-assembles context when the available budget changes by more than the refresh threshold (default 30%). When false, context is only assembled on explicit request."
}
},
"required": ["project", "view"],
"additionalProperties": false
}
Informal YAML Schema
The following annotated YAML provides an easier-to-read overview of the same schema defined formally above. It is not a validatable schema — it uses informal annotations to describe the structure. Refer to the JSON Schema above for the authoritative, machine-validatable definition.
# ─── Context View Identity ──────────────────────────────────────────
project: <namespace>/<name> # Project this context view applies to (required)
view: default | strategize | execute | apply # Which phase view (required)
# ─── Resource Filtering ─────────────────────────────────────────────
include_resources: # Resources to include (whitelist, optional)
- <resource_name>
exclude_resources: # Resources to exclude (blacklist, optional)
- <resource_name>
# ─── Path Filtering ─────────────────────────────────────────────────
include_paths: # Glob patterns for files to include (optional)
- "src/**/*.py"
- "tests/**/*.py"
exclude_paths: # Glob patterns for files to exclude (optional)
- "/node_modules/"
- "/pycache/"
- "/.git/"
- "/dist/"
# ─── Token and Size Budgets ─────────────────────────────────────────
hot_max_tokens: <integer> # Soft cap on hot context tokens (optional, null = no limit)
warm_max_decisions: <integer> # Max decisions in warm context (optional)
cold_max_decisions: <integer> # Max decisions in cold context (optional)
query_limit: <integer> # Max retrieval results per query (optional, default: 20)
max_file_size: <integer> # Max file size in bytes to include (optional, default: 1048576)
max_total_size: <integer> # Max total size across all included files (optional, default: 52428800)
# ─── Summarization ──────────────────────────────────────────────────
summarize: true # Enable summarization for large context segments (optional)
summary_max_tokens: <integer> # Token limit for generated summaries (optional)
# ─── ACMS Strategy & Context Assembly ───────────────────────────────
strategy: # ACMS strategies to use (optional, overrides global list)
- simple-keyword
- semantic-embedding
breadth-depth-navigator default_breadth: <integer> # Default hop count for UKO graph expansion (optional, default: 2) default_depth: 3 # Default detail depth — integer or named level (optional, default: 3) depth_gradient: # Per-hop detail depth overrides (optional) 0: 9 # Focus nodes get depth 9 (FULL_SOURCE for code) 1: 4 # 1-hop neighbors get depth 4 (SIGNATURES for code) 2: 0 # 2-hop neighbors get depth 0 (MODULE_LISTING for code) skeleton_ratio: <float> # Fraction of budget for inherited plan skeleton (optional, default: 0.15) temporal_scope: current | recent | all # Temporal scope for UKO node resolution (optional, default: current) auto_refresh: true # Auto re-assemble context on budget change (optional, default: true)
Structure Reference
Top-Level Fields
| Field | Type | Required | Description |
|---|---|---|---|
project |
string | Yes | Fully qualified project name this view applies to. |
view |
string | Yes | Phase view: default (fallback for all phases), strategize, execute, or apply. Phase-specific views override the default view. |
include_resources |
list | No | Whitelist of resource names to include. When specified, only these resources provide context. |
exclude_resources |
list | No | Blacklist of resource names to exclude. Applied after include_resources. |
include_paths |
list | No | Glob patterns for file paths to include. When specified, only matching files are included. |
exclude_paths |
list | No | Glob patterns for file paths to exclude. Applied after include_paths. |
hot_max_tokens |
integer | No | Soft cap on the number of tokens in hot (immediate) context. The effective limit is the lesser of this value and the actor's hard context window limit. When null, no soft cap is applied. |
warm_max_decisions |
integer | No | Maximum number of decisions retained in warm context (recent decisions available for reference). |
cold_max_decisions |
integer | No | Maximum number of decisions retained in cold context (older decisions queryable via search). |
query_limit |
integer | No | Maximum number of retrieval results per query against the cold tier. Default: 20. |
max_file_size |
integer | No | Maximum individual file size in bytes that will be included in context. Files exceeding this are either excluded or summarized. Default: 1 MB (1,048,576). |
max_total_size |
integer | No | Maximum aggregate size in bytes across all included files. Default: 50 MB (52,428,800). |
summarize |
boolean | No | When true, large context segments are automatically summarized rather than excluded. Default: false. |
summary_max_tokens |
integer | No | Maximum tokens for each generated summary. Only applies when summarize is true. |
strategy |
list | No | Ordered list of ACMS context strategies to use for this view. Overrides the global context.strategies.enabled list. Built-in strategies: simple-keyword, semantic-embedding, breadth-depth-navigator, arce, temporal-archaeology, plan-decision-context. |
default_breadth |
integer | No | Default number of hops from focus nodes in the UKO graph for context expansion. 0 = focus nodes only. Default: 2. |
default_depth |
integer or string | No | Default detail depth for context fragments. Accepts a non-negative integer or a named level string from the active domain's DetailLevelMap (e.g., SIGNATURES, FULL_SOURCE). Default: 3. |
depth_gradient |
object | No | Per-hop detail depth overrides. Keys are hop distances (0 = focus node), values are integers or named level strings. Hops not listed use default_depth. |
skeleton_ratio |
number | No | Fraction of context budget reserved for inherited plan skeleton context. Range: 0.0–1.0. Default: 0.15. |
temporal_scope |
string | No | Temporal scope for UKO node resolution: current (only isCurrent nodes), recent (current + warm retention window), all (include historical versions). Default: current. |
auto_refresh |
boolean | No | When true, ACMS automatically re-assembles context when the available budget changes by more than the refresh threshold. Default: true. |
Context Tiers
CleverAgents uses a three-tier context architecture:
| Tier | Description | Controlled By |
|---|---|---|
| Hot | Immediate context sent in the LLM prompt. Includes active files, recent changes, and current task description. | hot_max_tokens, include_paths, exclude_paths |
| Warm | Recent decisions and intermediate results. Available for reference but not always in the prompt. | warm_max_decisions |
| Cold | Historical decisions, full project index, and archived artifacts. Queryable via retrieval. | cold_max_decisions, query_limit |
View Resolution Order
When a plan enters a phase, the context view is resolved as follows:
- Check for a phase-specific view (e.g.,
strategize). - Fall back to the
defaultview. - Fall back to system defaults (include all resources, no path filtering, default budgets).
Examples
Example 1: Basic Default View (Simple)
A default context view with minimal filtering:
# context-default.yaml # Apply: agents project context set --view default --include-path "src/**" \ # --exclude-path "**/node_modules/**" local/api-service # Or import this file and apply via the CLI.project: local/api-service view: default
exclude_paths:
- "/node_modules/"
- "/pycache/"
- "/.git/"
max_file_size: 1048576 # 1 MB
Example 2: Strategize-Optimized View (Medium)
A focused view for the Strategize phase that limits context to essential architecture files:
# context-strategize.yamlproject: local/api-service view: strategize
include_resources:
- local/api-repo
exclude_resources:
- local/staging-db
include_paths:
- "src/**/*.py"
- "docs/architecture/**"
- "README.md"
- "pyproject.toml"
exclude_paths:
- "/node_modules/"
- "/test_fixtures/"
- "/pycache/"
- "/migrations/"
hot_max_tokens: 12000 warm_max_decisions: 50 cold_max_decisions: 200 query_limit: 20
max_file_size: 524288 # 512 KB max_total_size: 10485760 # 10 MB
summarize: true summary_max_tokens: 800
This view keeps the Strategize phase focused on source code and architecture docs, limits hot context to 12K tokens, and enables summarization for files that exceed the size limits.
Example 3: Execute-Phase View with Full Resources (Medium)
An execution view with broader access for implementing changes:
# context-execute.yamlproject: local/api-service view: execute
include_resources:
- local/api-repo
- local/staging-db
include_paths:
- "src/**"
- "tests/**"
- "config/**"
- "scripts/**"
- "Makefile"
- "pyproject.toml"
- "requirements*.txt"
exclude_paths:
- "/node_modules/"
- "/pycache/"
- "/dist/"
- "**/*.pyc"
hot_max_tokens: 24000 warm_max_decisions: 100 cold_max_decisions: 500 query_limit: 30
max_file_size: 2097152 # 2 MB max_total_size: 104857600 # 100 MB
summarize: true summary_max_tokens: 1200
Example 4: Apply-Phase Minimal View (Simple)
A minimal view for the Apply phase, focused only on changed files and validation:
# context-apply.yamlproject: local/api-service view: apply
include_paths:
- "src/**"
- "tests/**"
exclude_paths:
- "/pycache/"
hot_max_tokens: 8000 warm_max_decisions: 20 cold_max_decisions: 50
summarize: false
Example 5: Large Monorepo View with Aggressive Filtering (Complex)
A context view designed for a large monorepo where careful filtering is essential:
# context-monorepo-strategize.yamlproject: local/platform view: strategize
include_resources:
- local/platform-repo
include_paths:
- "packages/auth/**/*.ts"
- "packages/auth/**/*.tsx"
- "packages/shared/**/*.ts"
- "packages/api-gateway/**/*.ts"
- "docs/architecture/**"
- "package.json"
- "tsconfig.json"
- "lerna.json"
exclude_paths:
- "/node_modules/"
- "/dist/"
- "/coverage/"
- "**/*.test.ts"
- "**/*.spec.ts"
- "**/*.stories.tsx"
- "/fixtures/"
- "/snapshots/"
- "/generated/"
hot_max_tokens: 16000 warm_max_decisions: 30 cold_max_decisions: 150 query_limit: 15
max_file_size: 262144 # 256 KB - aggressive for a monorepo max_total_size: 5242880 # 5 MB
summarize: true summary_max_tokens: 600
This configuration is tailored for a TypeScript monorepo, focusing on specific packages and excluding generated code, tests, and build artifacts to keep context within manageable bounds.
Automation Profile Configuration Files
!!! adr "Architecture Decision" The automation profile configuration schema, threshold model, and built-in profiles are defined in ADR-017: Automation Profiles.
Automation profile configuration files define named collections of confidence thresholds (floating-point values from 0.0 to 1.0) and boolean safety flags controlling which operations are automated vs. requiring human approval. Each threshold specifies the minimum confidence score at which the system proceeds automatically; below the threshold, the system drops to manual mode. A threshold of 0.0 means "always automatic" and 1.0 means "always manual." Built-in profiles (manual, review, supervised, cautious, trusted, auto, ci, full-auto) are always available. Custom profiles are registered via agents automation-profile add --config <file>.
JSON Schema
The following is the formal JSON Schema definition for automation profile configuration files.
{
"$schema": "https://json-schema.org/draft/2020-12/schema",
"$id": "https://cleveragents.dev/schemas/automation-profile-config.json",
"title": "CleverAgents Automation Profile Configuration",
"description": "Configuration file schema for defining automation profiles — named collections of confidence thresholds (0.0-1.0) and boolean safety flags controlling which operations are automated vs. requiring human approval.",
"type": "object",
"properties": {
"name": {
"type": "string",
"pattern": "^[a-zA-Z0-9_-]+/[a-zA-Z0-9_-]+$",
"description": "Fully qualified profile name in <namespace>/<name> format."
},
"description": {
"type": "string",
"description": "Human-readable description of the profile's purpose and behavior."
},
"auto_strategize": {
"type": "number",
"minimum": 0.0,
"maximum": 1.0,
"description": "Confidence threshold (0.0-1.0) for automatically transitioning from Action to Strategize. 0.0 = always automatic, 1.0 = always manual."
},
"auto_execute": {
"type": "number",
"minimum": 0.0,
"maximum": 1.0,
"description": "Confidence threshold (0.0-1.0) for automatically transitioning from Strategize to Execute. 0.0 = always automatic, 1.0 = always manual."
},
"auto_apply": {
"type": "number",
"minimum": 0.0,
"maximum": 1.0,
"description": "Confidence threshold (0.0-1.0) for automatically applying changes after execution completes. 0.0 = always automatic, 1.0 = always manual."
},
"auto_decisions_strategize": {
"type": "number",
"minimum": 0.0,
"maximum": 1.0,
"description": "Confidence threshold (0.0-1.0) for autonomously making decisions during Strategize. 0.0 = always automatic, 1.0 = always manual."
},
"auto_decisions_execute": {
"type": "number",
"minimum": 0.0,
"maximum": 1.0,
"description": "Confidence threshold (0.0-1.0) for autonomously making decisions during Execute. 0.0 = always automatic, 1.0 = always manual."
},
"auto_validation_fix": {
"type": "number",
"minimum": 0.0,
"maximum": 1.0,
"description": "Confidence threshold (0.0-1.0) for automatically attempting to fix validation failures. 0.0 = always automatic, 1.0 = always manual."
},
"auto_strategy_revision": {
"type": "number",
"minimum": 0.0,
"maximum": 1.0,
"description": "Confidence threshold (0.0-1.0) for automatically revising the strategy when execution reveals issues. 0.0 = always automatic, 1.0 = always manual."
},
"auto_reversion_from_apply": {
"type": "number",
"minimum": 0.0,
"maximum": 1.0,
"description": "Confidence threshold (0.0-1.0) for automatically reverting from a constrained Apply phase to Strategize. 0.0 = always automatic, 1.0 = always manual."
},
"auto_retry_transient": {
"type": "number",
"minimum": 0.0,
"maximum": 1.0,
"description": "Confidence threshold (0.0-1.0) for automatically retrying operations that fail due to transient errors. 0.0 = always automatic, 1.0 = always manual."
},
"auto_checkpoint_restore": {
"type": "number",
"minimum": 0.0,
"maximum": 1.0,
"description": "Confidence threshold (0.0-1.0) for automatically restoring from the most recent checkpoint on failure. 0.0 = always automatic, 1.0 = always manual."
},
"auto_child_plans": {
"type": "number",
"minimum": 0.0,
"maximum": 1.0,
"description": "Confidence threshold (0.0-1.0) for automatically spawning child plans decided during Strategize. 0.0 = always automatic, 1.0 = always manual."
},
"require_sandbox": {
"type": "boolean",
"description": "When true, all write operations must execute within a sandbox. Execution fails if no sandbox strategy is available."
},
"require_checkpoints": {
"type": "boolean",
"description": "When true, checkpoints must be created before any write operation, enabling rollback."
},
"allow_unsafe_tools": {
"type": "boolean",
"description": "When true, tools flagged as unsafe can be invoked. When false, unsafe tool invocations are blocked."
}
},
"required": [
"name",
"description",
"auto_strategize",
"auto_execute",
"auto_apply",
"auto_decisions_strategize",
"auto_decisions_execute",
"auto_validation_fix",
"auto_strategy_revision",
"auto_reversion_from_apply",
"auto_retry_transient",
"auto_checkpoint_restore",
"auto_child_plans",
"require_sandbox",
"require_checkpoints",
"allow_unsafe_tools"
],
"additionalProperties": false
}
Informal YAML Schema
The following annotated YAML provides an easier-to-read overview of the same schema defined formally above. It is not a validatable schema — it uses informal annotations to describe the structure. Refer to the JSON Schema above for the authoritative, machine-validatable definition.
# ─── Profile Identity ─────────────────────────────────────────────── name: <namespace>/<name> # Fully qualified profile name (required) description: <string> # Human-readable description (required)# ─── Phase Transition Thresholds ──────────────────────────────────── # Confidence thresholds controlling phase transitions. # 0.0 = always automatic, 1.0 = always manual. auto_strategize: <float: 0.0–1.0> # Confidence threshold for Action → Strategize (required, 0.0=auto, 1.0=manual) auto_execute: <float: 0.0–1.0> # Confidence threshold for Strategize → Execute (required, 0.0=auto, 1.0=manual) auto_apply: <float: 0.0–1.0> # Confidence threshold for Execute → Apply (required, 0.0=auto, 1.0=manual)
# ─── Decision Thresholds ──────────────────────────────────────────── # Confidence thresholds controlling decisions within each phase. # 0.0 = always automatic, 1.0 = always manual. auto_decisions_strategize: <float: 0.0–1.0> # Confidence threshold for decisions during Strategize (required, 0.0=auto, 1.0=manual) auto_decisions_execute: <float: 0.0–1.0> # Confidence threshold for decisions during Execute (required, 0.0=auto, 1.0=manual)
# ─── Self-Repair Thresholds ───────────────────────────────────────── auto_validation_fix: <float: 0.0–1.0> # Confidence threshold for auto-fixing validation failures (required, 0.0=auto, 1.0=manual) auto_strategy_revision: <float: 0.0–1.0> # Confidence threshold for auto-revising strategy when Execute hits constraints (required, 0.0=auto, 1.0=manual) auto_reversion_from_apply: <float: 0.0–1.0> # Confidence threshold for auto-reverting from constrained Apply to Strategize (required, 0.0=auto, 1.0=manual) auto_retry_transient: <float: 0.0–1.0> # Confidence threshold for auto-retrying transient errors (required, 0.0=auto, 1.0=manual) auto_checkpoint_restore: <float: 0.0–1.0> # Confidence threshold for auto-restoring from checkpoints (required, 0.0=auto, 1.0=manual)
# ─── Execution Control Thresholds ────────────────────────────────── auto_child_plans: <float: 0.0–1.0> # Confidence threshold for auto-spawning child plans (required, 0.0=auto, 1.0=manual)
# ─── Safety Requirements ──────────────────────────────────────────── require_sandbox: <boolean> # Require sandbox for all write operations (required) require_checkpoints: <boolean> # Require checkpoint creation before write operations (required) allow_unsafe_tools: <boolean> # Allow tools flagged as unsafe (required)
Structure Reference
Top-Level Fields
| Field | Type | Required | Description |
|---|---|---|---|
name |
string | Yes | Fully qualified profile name. This value is used as the profile's registered name. |
description |
string | Yes | Human-readable description of the profile's purpose and behavior. |
Phase Transition Thresholds
| Field | Type | Required | Description |
|---|---|---|---|
auto_strategize |
number (0.0–1.0) | Yes | Confidence threshold for transitioning from Action to Strategize. When computed confidence >= threshold, Strategize begins automatically. 0.0 = always automatic, 1.0 = always manual. |
auto_execute |
number (0.0–1.0) | Yes | Confidence threshold for transitioning from Strategize to Execute. When computed confidence >= threshold, Execute begins automatically. 0.0 = always automatic, 1.0 = always manual. |
auto_apply |
number (0.0–1.0) | Yes | Confidence threshold for transitioning from Execute to Apply. When computed confidence >= threshold, Apply begins automatically. 0.0 = always automatic, 1.0 = always manual. A threshold of 0.0 means changes are committed without human review. |
Decision Automation Thresholds
| Field | Type | Required | Description |
|---|---|---|---|
auto_decisions_strategize |
number (0.0–1.0) | Yes | Confidence threshold for autonomous decision-making during Strategize. When computed confidence >= threshold, the strategy actor decides autonomously. 0.0 = always automatic, 1.0 = always manual. |
auto_decisions_execute |
number (0.0–1.0) | Yes | Confidence threshold for autonomous decision-making during Execute. When computed confidence >= threshold, the execution actor decides autonomously. 0.0 = always automatic, 1.0 = always manual. |
Self-Repair Thresholds
| Field | Type | Required | Description |
|---|---|---|---|
auto_validation_fix |
number (0.0–1.0) | Yes | Confidence threshold for automatically fixing validation failures (e.g., failing tests, lint errors). When computed confidence >= threshold, fix is attempted automatically. 0.0 = always automatic, 1.0 = always manual. |
auto_strategy_revision |
number (0.0–1.0) | Yes | Confidence threshold for automatically revising the strategy when execution reveals issues. When computed confidence >= threshold, revision proceeds automatically. 0.0 = always automatic, 1.0 = always manual. |
auto_reversion_from_apply |
number (0.0–1.0) | Yes | Confidence threshold for automatically reverting from a constrained Apply phase to Strategize. When computed confidence >= threshold, reversion proceeds automatically. 0.0 = always automatic, 1.0 = always manual. |
auto_retry_transient |
number (0.0–1.0) | Yes | Confidence threshold for automatically retrying operations that fail due to transient errors (network timeouts, rate limits). When computed confidence >= threshold, retry proceeds automatically. 0.0 = always automatic, 1.0 = always manual. |
auto_checkpoint_restore |
number (0.0–1.0) | Yes | Confidence threshold for automatically restoring from the most recent checkpoint on failure. When computed confidence >= threshold, restore proceeds automatically. 0.0 = always automatic, 1.0 = always manual. |
Execution Control Thresholds
| Field | Type | Required | Description |
|---|---|---|---|
auto_child_plans |
number (0.0–1.0) | Yes | Confidence threshold for automatically spawning child plans decided during Strategize. When computed confidence >= threshold, child plans spawn automatically during Execute. 0.0 = always automatic, 1.0 = always manual. |
Safety Flags
| Field | Type | Required | Description |
|---|---|---|---|
require_sandbox |
boolean | Yes | When true, all write operations must execute within a sandbox (git worktree, copy-on-write, transaction). Execution fails if no sandbox strategy is available. |
require_checkpoints |
boolean | Yes | When true, checkpoints must be created before any write operation. This enables rollback. |
allow_unsafe_tools |
boolean | Yes | When true, tools flagged as unsafe can be invoked. When false, unsafe tool invocations are blocked. |
Built-in Profile Reference
| Threshold | manual |
review |
supervised |
cautious |
trusted |
auto |
ci |
full-auto |
|---|---|---|---|---|---|---|---|---|
auto_strat |
1.0 | 0.0 | 0.0 | 0.7 | 0.0 | 0.0 | 0.0 | 0.0 |
auto_exec |
1.0 | 0.0 | 1.0 | 0.7 | 0.0 | 0.0 | 0.0 | 0.0 |
auto_apply |
1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 0.0 | 0.0 |
auto_dec_strat |
1.0 | 1.0 | 0.0 | 0.6 | 0.0 | 0.0 | 0.0 | 0.0 |
auto_dec_exec |
1.0 | 1.0 | 1.0 | 0.8 | 0.0 | 0.0 | 0.0 | 0.0 |
auto_val_fix |
1.0 | 1.0 | 1.0 | 0.7 | 0.0 | 0.0 | 0.0 | 0.0 |
auto_strat_rev |
1.0 | 1.0 | 1.0 | 0.8 | 1.0 | 0.0 | 0.0 | 0.0 |
auto_rev_apply |
1.0 | 1.0 | 1.0 | 0.9 | 1.0 | 1.0 | 0.0 | 0.0 |
auto_child |
1.0 | 0.0 | 1.0 | 0.7 | 0.0 | 0.0 | 0.0 | 0.0 |
require_sandbox |
true | true | true | true | true | true | true | false |
allow_unsafe |
false | false | false | false | false | false | false | true |
Examples
Example 1: Careful Autonomous Profile (Simple)
A profile for autonomous execution with mandatory sandboxing and manual apply:
# careful-auto.yaml # Register: agents automation-profile add --config careful-auto.yamlname: local/careful-auto description: "Autonomous execution with mandatory sandbox and manual apply"
auto_strategize: 0.0 auto_execute: 0.0 auto_apply: 1.0
auto_decisions_strategize: 0.0 auto_decisions_execute: 0.0
auto_validation_fix: 0.0 auto_strategy_revision: 1.0 auto_reversion_from_apply: 1.0 auto_retry_transient: 0.0 auto_checkpoint_restore: 0.0
auto_child_plans: 0.0
require_sandbox: true require_checkpoints: true allow_unsafe_tools: false
This is essentially the auto built-in profile but with auto_checkpoint_restore set to 0.0 (always automatic) for more aggressive self-repair.
Example 2: CI/CD Pipeline Profile (Medium)
A profile designed for automated CI/CD pipelines where everything runs autonomously:
# ci-pipeline.yaml # Register: agents automation-profile add --config ci-pipeline.yamlname: local/ci-pipeline description: "Full automation for CI/CD pipelines. All phases automated, sandbox required."
auto_strategize: 0.0 auto_execute: 0.0 auto_apply: 0.0
auto_decisions_strategize: 0.0 auto_decisions_execute: 0.0
auto_validation_fix: 0.0 auto_strategy_revision: 0.0 auto_reversion_from_apply: 0.0 auto_retry_transient: 0.0 auto_checkpoint_restore: 0.0
auto_child_plans: 0.0
require_sandbox: true require_checkpoints: true allow_unsafe_tools: false
This profile is similar to full-auto but with mandatory sandboxing and checkpoints enabled — suitable for CI/CD environments where you want full automation but with safety nets.
Example 3: Review-Heavy Profile (Medium)
A profile for regulated environments where every decision must be reviewed:
# review-heavy.yaml # Register: agents automation-profile add --config review-heavy.yamlname: local/review-heavy description: "Maximum human oversight. Every decision and phase requires approval."
auto_strategize: 0.0 auto_execute: 1.0 auto_apply: 1.0
auto_decisions_strategize: 1.0 auto_decisions_execute: 1.0
auto_validation_fix: 1.0 auto_strategy_revision: 1.0 auto_reversion_from_apply: 1.0 auto_retry_transient: 1.0 auto_checkpoint_restore: 1.0
auto_child_plans: 1.0
require_sandbox: true require_checkpoints: true allow_unsafe_tools: false
This profile auto-starts strategize but requires human approval for every decision, every phase transition, and every child plan spawn. Useful in regulated industries (finance, healthcare) where audit trails and human review are mandatory.
Example 4: Development Sandbox Profile (Simple)
A fast-iteration profile for local development where safety is relaxed:
# dev-sandbox.yaml # Register: agents automation-profile add --config dev-sandbox.yamlname: local/dev-sandbox description: "Fast iteration for local development. Relaxed safety, auto execution."
auto_strategize: 0.0 auto_execute: 0.0 auto_apply: 1.0
auto_decisions_strategize: 0.0 auto_decisions_execute: 0.0
auto_validation_fix: 0.0 auto_strategy_revision: 0.0 auto_reversion_from_apply: 1.0 auto_retry_transient: 0.0 auto_checkpoint_restore: 0.0
auto_child_plans: 0.0
require_sandbox: false require_checkpoints: false allow_unsafe_tools: true
This profile removes sandbox and checkpoint requirements for speed during local development. It still requires manual apply to prevent accidental commits. Unsafe tools are allowed for full flexibility.
Example 5: Production Deployment Profile (Complex)
A profile for production deployments combining autonomous execution with maximum safety:
# production-deploy.yaml # Register: agents automation-profile add --config production-deploy.yamlname: local/production-deploy description: | Production deployment profile. Fully autonomous execution with maximum safety guarantees. All phases require sandbox and checkpoint. Apply requires manual approval. Strategy revision is enabled to adapt to deployment issues.
auto_strategize: 0.0 auto_execute: 0.0 auto_apply: 1.0
auto_decisions_strategize: 0.0 auto_decisions_execute: 0.0
auto_validation_fix: 0.0 auto_strategy_revision: 0.0 auto_reversion_from_apply: 1.0 auto_retry_transient: 0.0 auto_checkpoint_restore: 0.0
auto_child_plans: 0.0
require_sandbox: true require_checkpoints: true allow_unsafe_tools: false
Workflow Examples
This section presents complete, end-to-end workflow examples showing how CleverAgents is used in practice. Each example starts from initial setup and walks through every command and decision point. The examples progress from simple single-action workflows to complex multi-project, multi-actor orchestrations.
Example 1: Hello World — Fix a Single Bug
Scenario: A developer finds a bug in their Python web API where the /health endpoint returns a 500 error instead of 200 when the database connection is unavailable. They want CleverAgents to fix it with full human oversight.
Complexity: Beginner. Single project, single action, manual automation profile.
Step 1: Initialize Environment and Register Resources
# Initialize CleverAgents (first-time setup) $ agents init --yes╭─ Initialized ────────────────────────────────────────────╮ │ Data Dir: /home/dev/.cleveragents (created) │ │ Config: /home/dev/.cleveragents/config.toml │ │ Database: initialized (schema v3) │ │ Directories: logs, cache, sessions, contexts │ ╰──────────────────────────────────────────────────────────╯
✓ OK Initialized (non-interactive)
# Register the git checkout as a resource $ agents resource add git-checkout local/api-repo \ --path /home/dev/projects/api-service \ --branch main╭─ Resource ──────────────────────────────────────────╮ │ Name: local/api-repo │ │ ID: 01HXR1A1B2C3D4E5F6G7H8J9K0 │ │ Type: git-checkout │ │ Physical/Virtual: physical │ │ Path: /home/dev/projects/api-service │ │ Branch: main │ │ Created: 2026-02-11 09:15 │ ╰─────────────────────────────────────────────────────╯
╭─ Auto-discovered Children ─────────────────────────────────────────╮ │ ID Type Status │ │ ─────────────── ────────────── ───────────────── │ │ 01HXR1A1B2C3… git created │ │ 01HXR1A1B2C4… git-remote created │ │ 01HXR1A1B2C5… git-branch created │ │ 01HXR1A1B2C6… fs-directory created │ │ + 32 git-commit resources │ │ + 187 git-tree-entry resources │ │ + 5 fs-directory + 41 fs-file │ ╰────────────────────────────────────────────────────────────────────╯
╭─ Capabilities ─────────────────╮ │ Readable: yes │ │ Writable: yes │ │ Sandboxable: yes │ │ Checkpointable: yes │ │ Sandbox Strategy: git_worktree │ ╰────────────────────────────────╯
✓ OK Resource registered (270 child resources discovered)
# Create the project and link the resource $ agents project create \ --description "REST API service" \ --resource local/api-repo \ local/api-service╭─ Project Created ──────────────────────────╮ │ Name: local/api-service │ │ Description: REST API service │ │ Type: local │ │ Created: 2026-02-11 09:15 │ ╰────────────────────────────────────────────╯
╭─ Linked Resources ───────────────────────────────────────╮ │ Name Type Read-Only │ │ ───────────────── ────────────── ───────── │ │ local/api-repo git-checkout no │ ╰──────────────────────────────────────────────────────────╯
╭─ Defaults ──────────────────────────────╮ │ Sandbox: git_worktree │ │ Validations: 0 │ │ Context Filters: none │ │ Automation Profile: (inherits global) │ ╰─────────────────────────────────────────╯
✓ OK Project created
# Register a required validation (tests must pass) $ agents validation add \ --config validations/unit-tests.yaml \ --required \ local/unit-tests╭─ Validation Registered ─────────────────────────────────────╮ │ Name: local/unit-tests │ │ Command: pytest tests/ │ │ Description: Unit tests │ │ Mode: required │ │ Timeout: 300s (default) │ ╰─────────────────────────────────────────────────────────────╯
✓ OK Validation registered
# Attach the validation to the project $ agents validation attach --project local/api-service local/unit-tests
✓ OK Validation attached to project local/api-service
Step 2: Create an Action
Create a file actions/fix-bug.yaml:
name: local/fix-bug
description: "Fix a reported bug in the codebase"
long_description: |
Analyze the bug description, locate the relevant code,
implement a fix, and add or update tests to cover the fix.
definition_of_done: |
- The described bug no longer reproduces
- All existing tests pass
- At least one new test covers the fixed behavior
- No unrelated code is modified
strategy_actor: anthropic/claude-3.5-sonnet
execution_actor: anthropic/claude-3.5-sonnet
reusable: true
state: available
args:
- name: bug_description
type: string
required: true
description: "Description of the bug to fix"
- name: affected_file
type: string
required: false
description: "File path where the bug is suspected (optional hint)"
# Register the action $ agents action create --config ./actions/fix-bug.yaml╭─ Action Created ─────────────────────────────╮ │ Name: local/fix-bug │ │ ID: 01HXR1B1C2D3E4F5G6H7I8J9K0 │ │ State: available │ │ Strategy Actor: anthropic/claude-3.5-sonnet │ │ Execution Actor: anthropic/claude-3.5-sonnet │ │ Reusable: yes │ │ Read Only: no │ │ Created: 2026-02-11 09:16 │ ╰──────────────────────────────────────────────╯
╭─ Definition of Done ──────────────────────────────────╮ │ - The described bug no longer reproduces │ │ - All existing tests pass │ │ - At least one new test covers the fixed behavior │ │ - No unrelated code is modified │ ╰───────────────────────────────────────────────────────╯
╭─ Arguments ───────────────────────────────────────────────────────────────╮ │ Name Type Required Description │ │ ─────────────── ────── ──────── ────────────────────────────────────── │ │ bug_description string yes Description of the bug to fix │ │ affected_file string no File path where the bug is suspected │ ╰───────────────────────────────────────────────────────────────────────────╯
╭─ Automation ──────────────────────────╮ │ Profile: (inherits global) │ │ Source: default │ ╰───────────────────────────────────────╯
╭─ Usage ──────────────────────────────────────────────────────────────────╮ │ agents plan use local/fix-bug local/api-service │ │ --arg bug_description="..." │ ╰──────────────────────────────────────────────────────────────────────────╯
✓ OK Action created
Step 3: Create and Execute a Plan
# Bind the action to the project with the manual profile for full oversight $ agents plan use \ --automation-profile manual \ --arg bug_description="The /health endpoint returns HTTP 500 when the database \ connection is unavailable. It should return HTTP 200 with {\"status\": \"degraded\", \ \"database\": \"unavailable\"} instead of crashing." \ --arg affected_file="src/routes/health.py" \ local/fix-bug local/api-service╭─ Plan Created ──────────────────────────╮ │ Plan ID: 01HXR1C1D2E3F4G5H6I7J8K9L0 │ │ Phase: strategize │ │ Action: local/fix-bug │ │ Project: local/api-service │ │ Automation: manual │ │ Attempt: 1 │ ╰─────────────────────────────────────────╯
╭─ Inputs ──────────────────────────────────────────────────────────────╮ │ - bug_description="The /health endpoint returns HTTP 500 when the │ │ database connection is unavailable. It should return HTTP 200 with │ │ {"status": "degraded", "database": "unavailable"} instead │ │ of crashing." │ │ - affected_file="src/routes/health.py" │ │ - automation_profile=manual │ ╰───────────────────────────────────────────────────────────────────────╯
╭─ Actors ──────────────────────────────────────╮ │ Strategy: anthropic/claude-3.5-sonnet │ │ Execution: anthropic/claude-3.5-sonnet │ │ Estimation: (none) │ ╰───────────────────────────────────────────────╯
╭─ Automation ──────────────────────────╮ │ Profile: manual │ │ Source: plan override │ │ Read-Only: no │ ╰───────────────────────────────────────╯
╭─ Context ─────────────────────────╮ │ Resources: 1 (git-checkout) │ │ Indexed Files: 41 │ │ View: strategize │ │ Hot Token Budget: 8,000 │ ╰───────────────────────────────────╯
╭─ Next Steps ───────────────────────────────────────╮ │ - agents plan execute 01HXR1C1D2E3F4G5H6I7J8K9L0 │ │ - agents plan status 01HXR1C1D2E3F4G5H6I7J8K9L0 │ │ - agents plan tree 01HXR1C1D2E3F4G5H6I7J8K9L0 │ ╰────────────────────────────────────────────────────╯
✓ OK Plan created
Because the automation profile is manual, the plan stops and waits after creation. The developer must explicitly trigger each phase.
# Manually trigger Strategize → the strategy actor analyzes the codebase $ agents plan execute 01HXR1C1D2E3F4G5H6I7J8K9L0╭─ Execution ─────────────────────────────╮ │ Plan: 01HXR1C1D2E3F4G5H6I7J8K9L0 │ │ Phase: strategize │ │ Sandbox: git_worktree │ │ Worker: anthropic/claude-3.5-sonnet │ │ Started: 09:17:02 │ │ Attempt: 1 │ ╰─────────────────────────────────────────╯
╭─ Sandbox ───────────────────────────────────────────────────────╮ │ Strategy: git_worktree │ │ Path: /home/dev/projects/api-service/.worktrees/plan-01HXR1C1 │ │ Branch: cleveragents/plan-01HXR1C1D2 │ │ Status: active │ ╰─────────────────────────────────────────────────────────────────╯
╭─ Strategy Summary ────────────────╮ │ Decisions: 4 │ │ Invariants: 0 │ │ Planned Child Plans: 0 │ │ Estimated Files: ~2 │ │ Risk: low │ ╰───────────────────────────────────╯
╭─ Progress ────────╮ │ ✓ Collect context │ │ ✓ Build strategy │ │ • Awaiting review │ ╰───────────────────╯
✓ OK Strategize complete — awaiting manual approval
The strategy actor has read the codebase, identified the relevant files, and produced decisions. Since auto_decisions_strategize is 1.0 (manual), the plan pauses for review:
# Review the decision tree $ agents plan tree 01HXR1C1D2E3F4G5H6I7J8K9L0╭─ Decision Tree ───────────────────────────────────────────────────────────────────────────╮ │ Plan: 01HXR1C1D2E3F4G5H6I7J8K9L0 │ │ ├─ [prompt_definition] "Fix /health endpoint to return 200 when DB unavailable" │ │ ├─ [strategy_choice] "Wrap DB call in try/except, return degraded status" (conf: 0.95) │ │ ├─ [strategy_choice] "Modify src/routes/health.py only" (conf: 0.91) │ │ └─ [strategy_choice] "Add test_health_db_unavailable to tests/test_health.py" (conf: 0.93)│ ╰───────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Tree Summary ─────────────╮ │ Nodes: 4 │ │ Depth: 1 │ │ Child Plans: 0 │ │ Invariants: 0 │ │ Superseded: 0 (hidden) │ ╰────────────────────────────╯
╭─ Decision IDs (for correction) ──────────────────╮ │ Root: 01HXR1D1A2B3C4D5E6F7G8H9 │ │ Strategy 1: 01HXR1D2A2B3C4D5E6F7G8H9 │ │ Strategy 2: 01HXR1D3A2B3C4D5E6F7G8H9 │ │ Strategy 3: 01HXR1D4A2B3C4D5E6F7G8H9 │ ╰──────────────────────────────────────────────────╯
✓ OK Decision tree rendered
# Inspect the main strategy decision in detail $ agents plan explain 01HXR1D2A2B3C4D5E6F7G8H9╭─ Decision ──────────────────────────────────────────────────────────────╮ │ ID: 01HXR1D2A2B3C4D5E6F7G8H9 │ │ Type: strategy_choice │ │ Question: How should the /health endpoint handle database failures? │ │ Chosen: Wrap DB call in try/except, return degraded status │ │ Confidence: 0.95 │ │ Plan: 01HXR1C1D2E3F4G5H6I7J8K9L0 │ │ Sequence: 1 of 3 │ │ Created: 2026-02-11 09:17 │ ╰─────────────────────────────────────────────────────────────────────────╯
╭─ Alternatives Considered ──────────────────────────────────────────────────╮ │ 1. Wrap DB call in try/except, return degraded status (chosen) │ │ 2. Add a circuit breaker middleware for all DB-dependent routes │ │ 3. Return HTTP 503 Service Unavailable with retry-after header │ ╰────────────────────────────────────────────────────────────────────────────╯
╭─ Impact ──────────────────────╮ │ Downstream Decisions: 2 │ │ Downstream Child Plans: 0 │ │ Artifacts Produced: 2 │ │ Correction Impact: low │ ╰───────────────────────────────╯
╭─ Rationale ──────────────────────────────────────────────────╮ │ The bug description explicitly requires HTTP 200 with a │ │ degraded status JSON body. A try/except around the DB call │ │ in the health endpoint is the minimal, targeted fix. A │ │ circuit breaker would be over-engineering for a single │ │ endpoint. HTTP 503 contradicts the requested behavior. │ ╰──────────────────────────────────────────────────────────────╯
╭─ Correction ─────────────────────────────────────────────────╮ │ agents plan correct 01HXR1D2A2B3C4D5E6F7G8H9 │ │ --mode revert --guidance "Use a different approach..." │ ╰──────────────────────────────────────────────────────────────╯
✓ OK Decision explained
# Approve and proceed to Execute (manual profile requires explicit trigger) $ agents plan execute 01HXR1C1D2E3F4G5H6I7J8K9L0╭─ Execution ─────────────────────────────╮ │ Plan: 01HXR1C1D2E3F4G5H6I7J8K9L0 │ │ Phase: execute │ │ Sandbox: git_worktree │ │ Worker: anthropic/claude-3.5-sonnet │ │ Started: 09:18:30 │ │ Attempt: 1 │ ╰─────────────────────────────────────────╯
╭─ Sandbox ───────────────────────────────────────────────────────╮ │ Strategy: git_worktree │ │ Path: /home/dev/projects/api-service/.worktrees/plan-01HXR1C1 │ │ Branch: cleveragents/plan-01HXR1C1D2 │ │ Status: active │ ╰─────────────────────────────────────────────────────────────────╯
╭─ Progress ────────────────────────────────╮ │ ✓ Collect context (0.4s) │ │ ✓ Modify src/routes/health.py (1.2s) │ │ ✓ Add tests/test_health.py test (2.1s) │ │ ✓ Validate: pytest tests/ (3.8s) │ ╰───────────────────────────────────────────╯
✓ OK Execution complete — awaiting apply
The execution actor worked in the sandbox (git worktree), modifying src/routes/health.py and adding a test in tests/test_health.py. Validations ran automatically within the sandbox and passed.
Step 4: Review and Apply
# Review the changes $ agents plan diff 01HXR1C1D2E3F4G5H6I7J8K9L0╭─ Diff Summary ──────────────────────────────────╮ │ Plan: 01HXR1C1D2E3F4G5H6I7J8K9L0 │ │ Project: local/api-service │ │ Files Changed: 2 │ │ Insertions: 19 │ │ Deletions: 3 │ │ Net Change: +16 lines │ ╰─────────────────────────────────────────────────╯
╭─ Files ────────────────────────────────────╮ │ Path Change Status │ │ ─────────────────────── ────── ──────── │ │ src/routes/health.py +8 -3 modified │ │ tests/test_health.py +11 -0 modified │ ╰────────────────────────────────────────────╯
╭─ Patch Preview ──────────────────────────────────────────────────╮ │ --- a/src/routes/health.py │ │ +++ b/src/routes/health.py │ │ @@ -8,6 +8,11 @@ │ │ @app.get("/health") │ │ def health_check(): │ │ - db_status = check_database_connection() │ │ - return {"status": "ok", "database": db_status} │ │ + try: │ │ + db_status = check_database_connection() │ │ + return {"status": "ok", "database": "available"} │ │ + except ConnectionError: │ │ + return {"status": "degraded", "database": "unavailable"}│ │ --- a/tests/test_health.py │ │ +++ b/tests/test_health.py │ │ @@ -15,0 +16,11 @@ │ │ +def test_health_db_unavailable(client, monkeypatch): │ │ + def mock_check(): │ │ + raise ConnectionError("DB unreachable") │ │ + monkeypatch.setattr("src.routes.health.check_database…") │ │ + response = client.get("/health") │ │ + assert response.status_code == 200 │ │ + data = response.json() │ │ + assert data["status"] == "degraded" │ │ + assert data["database"] == "unavailable" │ ╰──────────────────────────────────────────────────────────────────╯
╭─ Risk Assessment ────────────────╮ │ API Compatibility: preserved │ │ Test Coverage: increased │ │ Breaking Changes: none detected │ ╰──────────────────────────────────╯
✓ OK Diff generated
# Check plan status (should be execute/complete, awaiting apply) $ agents plan status 01HXR1C1D2E3F4G5H6I7J8K9L0╭─ Plan Status ───────────────────────────╮ │ Plan: 01HXR1C1D2E3F4G5H6I7J8K9L0 │ │ Phase: execute │ │ State: complete │ │ Action: local/fix-bug │ │ Project: local/api-service │ │ Automation: manual │ │ Attempt: 1 │ ╰─────────────────────────────────────────╯
╭─ Progress ───────────╮ │ ✓ Strategize │ │ ✓ Execute │ │ • Apply (awaiting) │ ╰──────────────────────╯
╭─ Timing ───────────────────────────╮ │ Started: 2026-02-11 09:17:02 │ │ Elapsed: 00:01:42 │ │ ETA: awaiting manual apply │ ╰────────────────────────────────────╯
╭─ Execution Detail ──────────╮ │ Sandbox: git_worktree │ │ Tool Calls: 5 │ │ Files Modified: 2 │ │ Child Plans: 0 │ │ Checkpoints: 1 created │ ╰─────────────────────────────╯
╭─ Cost ───────────────╮ │ Tokens Used: 6,840 │ │ Cost So Far: $0.023 │ │ Estimated: $0.025 │ ╰──────────────────────╯
✓ OK Status refreshed
# Apply the changes to the real repository $ agents plan apply --yes 01HXR1C1D2E3F4G5H6I7J8K9L0╭─ Apply Summary ──────────────────────────╮ │ Plan: 01HXR1C1D2E3F4G5H6I7J8K9L0 │ │ Artifacts: 2 files updated │ │ Changes: 19 insertions, 3 deletions │ │ Project: local/api-service │ │ Applied At: 2026-02-11 09:19 │ ╰──────────────────────────────────────────╯
╭─ Validation ───────────────────╮ │ Tests: passed (14/14) │ │ Duration: 3.8s │ ╰────────────────────────────────╯
╭─ Sandbox Cleanup ──────────────────────╮ │ Worktree: removed │ │ Branch: merged to main │ │ Checkpoint: archived │ ╰────────────────────────────────────────╯
╭─ Plan Lifecycle ──────────────────────╮ │ Phase: apply │ │ State: applied │ │ Total Duration: 00:01:58 │ │ Total Cost: $0.025 │ │ Decisions Made: 4 │ │ Child Plans: 0 │ ╰───────────────────────────────────────╯
╭─ Next Steps ──────╮ │ - Review git diff │ │ - Commit changes │ ╰───────────────────╯
✓ OK Changes applied
The sandbox changes are merged into the main working tree. The bug is fixed.
Example 2: Automated Test Generation for a Module
Scenario: A team wants to increase test coverage for their auth module from 45% to 80%. They trust CleverAgents to generate tests autonomously, but want to review before applying.
Complexity: Intermediate. Single project, trusted automation profile, validation-driven.
Step 1: Setup (Assuming Project Exists)
# Register a coverage validation # (project local/api-service and resource local/api-service-repo already exist) $ agents validation add \ --config validations/auth-coverage.yaml \ --required \ local/auth-coverage╭─ Validation Registered ──────────────────────────────────────╮ │ Name: local/auth-coverage │ │ Command: pytest --cov=src/auth --cov-fail-under=80 │ │ tests/test_auth/ │ │ Description: Coverage must be at least 80% │ │ Mode: required │ │ Timeout: 600s │ ╰──────────────────────────────────────────────────────────────╯
✓ OK Validation registered
$ agents validation attach --project local/api-service local/auth-coverage
✓ OK Validation attached to project local/api-service
# Automation profile will be set per-plan via --automation-profile on plan use
Step 2: Create the Test Generation Action
Create actions/generate-tests.yaml:
name: local/generate-tests
description: "Generate comprehensive tests for a module"
long_description: |
Analyze the target module's code paths, identify untested branches,
edge cases, and error conditions, then generate thorough test cases.
Uses the project's existing test framework and conventions.
definition_of_done: |
- Coverage of the target module reaches the specified threshold
- All new tests pass
- All existing tests continue to pass
- Tests follow existing project conventions (fixtures, naming, structure)
- No production code is modified
strategy_actor: anthropic/claude-3.5-sonnet
execution_actor: anthropic/claude-3.5-sonnet
reusable: true
state: available
read_only: false
args:
- name: target_module
type: string
required: true
description: "Module path to generate tests for (e.g., src/auth)"
- name: coverage_target
type: integer
required: false
description: "Target coverage percentage"
default: 80
invariants:
- "Do not modify any production code — only add or modify test files"
- "Follow the existing test file naming convention (test_<module>.py)"
- "Use the project's existing test fixtures and conftest.py patterns"
$ agents action create --config ./actions/generate-tests.yaml╭─ Action Created ───────────────────────────────────╮ │ Name: local/generate-tests │ │ ID: 01HXR2B1C2D3E4F5G6H7I8J9K0 │ │ State: available │ │ Strategy Actor: anthropic/claude-3.5-sonnet │ │ Execution Actor: anthropic/claude-3.5-sonnet │ │ Reusable: yes │ │ Read Only: no │ │ Config: ./actions/generate-tests.yaml │ │ Created: 2026-02-11 09:15 │ ╰────────────────────────────────────────────────────╯
╭─ Definition of Done ──────────────────────────────────────────────╮ │ - Coverage of the target module reaches the specified threshold │ │ - All new tests pass │ │ - All existing tests continue to pass │ │ - Tests follow existing project conventions │ │ - No production code is modified │ ╰───────────────────────────────────────────────────────────────────╯
╭─ Arguments ───────────────────────────────────────────────────────────────────╮ │ Name Type Required Description │ │ ──────────────── ─────── ──────── ─────────────────────────────────────── │ │ target_module string yes Module path to generate tests for │ │ coverage_target integer no Target coverage percentage (default 80) │ ╰───────────────────────────────────────────────────────────────────────────────╯
╭─ Invariants ──────────────────────────────────────────────────────────────────╮ │ 1. Do not modify any production code — only add or modify test files │ │ 2. Follow the existing test file naming convention (test_<module>.py) │ │ 3. Use the project's existing test fixtures and conftest.py patterns │ ╰───────────────────────────────────────────────────────────────────────────────╯
╭─ Automation ──────────────────────────╮ │ Profile: supervised │ │ Source: default │ ╰───────────────────────────────────────╯
✓ OK Action created
Step 3: Execute
# With trusted profile: Strategize and Execute run automatically, # only Apply requires human approval $ agents plan use \ --automation-profile trusted \ --arg target_module="src/auth" \ --arg coverage_target=80 \ local/generate-tests local/api-service╭─ Plan Created ──────────────────────────────────────╮ │ Plan: 01HXR2B3C4D5E6F7G8H9J0K1L2 │ │ Phase: strategize │ │ State: processing │ │ Action: local/generate-tests │ │ Project: local/api-service │ │ Automation: trusted │ │ Attempt: 1 │ ╰─────────────────────────────────────────────────────╯
╭─ Inputs ────────────────────────────╮ │ - target_module=src/auth │ │ - coverage_target=80 │ │ - automation_profile=trusted │ ╰─────────────────────────────────────╯
╭─ Actors ────────────────────────────────────────╮ │ Strategy: anthropic/claude-3.5-sonnet │ │ Execution: anthropic/claude-3.5-sonnet │ │ Estimation: (none) │ ╰─────────────────────────────────────────────────╯
╭─ Automation ───────────────────────────╮ │ Profile: trusted │ │ Source: plan override │ │ Read-Only: no │ ╰────────────────────────────────────────╯
╭─ Context ────────────────────────╮ │ Resources: 1 (repo) │ │ Indexed Files: 214 │ │ View: strategize │ │ Hot Token Budget: 8,000 │ ╰──────────────────────────────────╯
╭─ Next Steps ─────────────────────────────────────────────────────────╮ │ Trusted profile — strategize and execute will proceed automatically. │ │ - agents plan status 01HXR2B3C4D5E6F7G8H9J0K1L2 │ │ - agents plan tree 01HXR2B3C4D5E6F7G8H9J0K1L2 │ ╰──────────────────────────────────────────────────────────────────────╯
✓ OK Plan created
The system runs through Strategize and Execute automatically. The strategy actor analyzes coverage gaps, identifies 12 untested code paths, and plans test cases. The execution actor writes tests in the sandbox. Validations run. If coverage reaches 80%, the plan reaches execute/complete and waits for apply.
# Check progress while it runs $ agents plan status 01HXR2B3C4D5E6F7G8H9J0K1L2╭─ Plan Status ───────────────────────────────╮ │ Plan: 01HXR2B3C4D5E6F7G8H9J0K1L2 │ │ Phase: execute │ │ State: processing │ │ Action: local/generate-tests │ │ Project: local/api-service │ │ Automation: trusted │ │ Attempt: 1 │ ╰─────────────────────────────────────────────╯
╭─ Progress ───────╮ │ ✓ Strategize │ │ ⏳ Execute │ │ • Apply (queued) │ ╰──────────────────╯
╭─ Timing ──────────╮ │ Started: 09:15:42 │ │ Elapsed: 00:02:18 │ │ ETA: 00:04:30 │ ╰───────────────────╯
╭─ Execution Detail ──────────╮ │ Sandbox: git_worktree │ │ Tool Calls: 14 │ │ Files Modified: 5 │ │ Child Plans: 0 │ │ Checkpoints: 3 created │ ╰─────────────────────────────╯
╭─ Cost ───────────────╮ │ Tokens Used: 18,740 │ │ Cost So Far: $0.062 │ │ Estimated: $0.110 │ ╰──────────────────────╯
✓ OK Status refreshed
# When complete, review the generated tests $ agents plan diff 01HXR2B3C4D5E6F7G8H9J0K1L2╭─ Diff Summary ─────────────────────────────────────╮ │ Plan: 01HXR2B3C4D5E6F7G8H9J0K1L2 │ │ Project: local/api-service │ │ Files Changed: 5 │ │ Insertions: 287 │ │ Deletions: 3 │ │ Net Change: +284 lines │ ╰────────────────────────────────────────────────────╯
╭─ Files ────────────────────────────────────────────────╮ │ Path Change Status │ │ ────────────────────────────── ──────── ────────── │ │ tests/test_auth/test_login.py +72 -0 new file │ │ tests/test_auth/test_tokens.py +68 -0 new file │ │ tests/test_auth/test_session.py +54 -0 new file │ │ tests/test_auth/test_rbac.py +41 -0 new file │ │ tests/test_auth/conftest.py +52 -3 modified │ ╰────────────────────────────────────────────────────────╯
╭─ Patch Preview ───────────────────────────────────────────╮ │ --- /dev/null │ │ +++ b/tests/test_auth/test_login.py │ │ @@ -0,0 +1,72 @@ │ │ + import pytest │ │ + from src.auth.login import authenticate, verify_mfa │ │ + ... │ │ --- a/tests/test_auth/conftest.py │ │ +++ b/tests/test_auth/conftest.py │ │ @@ -1,5 +1,54 @@ │ │ - # minimal fixtures │ │ + # shared auth test fixtures │ │ + @pytest.fixture │ │ + def mock_user_db(): ... │ ╰───────────────────────────────────────────────────────────╯
╭─ Risk Assessment ────────────────╮ │ API Compatibility: preserved │ │ Test Coverage: 45% → 83% │ │ Breaking Changes: none detected │ ╰──────────────────────────────────╯
✓ OK Diff generated
$ agents plan artifacts 01HXR2B3C4D5E6F7G8H9J0K1L2╭─ Artifacts ──────────────────────────────────────────────────────────╮ │ Path Type Size Change Child Plan │ │ ──────────────────────────────── ───── ────── ────── ─────── │ │ tests/test_auth/test_login.py write 3.8 KB +72 -0 root │ │ tests/test_auth/test_tokens.py write 3.4 KB +68 -0 root │ │ tests/test_auth/test_session.py write 2.7 KB +54 -0 root │ │ tests/test_auth/test_rbac.py write 2.1 KB +41 -0 root │ │ tests/test_auth/conftest.py edit 2.6 KB +52 -3 root │ ╰──────────────────────────────────────────────────────────────────────╯
╭─ Summary ────────────╮ │ Total: 5 │ │ Writes: 4 (new) │ │ Edits: 1 (modified) │ │ Deletes: 0 │ │ Total Size: 14.6 KB │ ╰──────────────────────╯
✓ OK 5 artifacts listed
# Apply the generated tests to the real repository $ agents plan apply --yes 01HXR2B3C4D5E6F7G8H9J0K1L2╭─ Apply Summary ──────────────────────────╮ │ Plan: 01HXR2B3C4D5E6F7G8H9J0K1L2 │ │ Artifacts: 5 files updated │ │ Changes: 287 insertions, 3 deletions │ │ Project: local/api-service │ │ Applied At: 2026-02-11 09:22 │ ╰──────────────────────────────────────────╯
╭─ Validation ──────────────────────────╮ │ ✓ Coverage: passed (83% ≥ 80%) │ │ ✓ All tests: passed (47/47) │ │ Duration: 18.2s │ ╰───────────────────────────────────────╯
╭─ Sandbox Cleanup ─────────╮ │ Worktree: removed │ │ Branch: merged to main │ │ Checkpoint: archived │ ╰───────────────────────────╯
╭─ Plan Lifecycle ─────────────────────────╮ │ Phase: apply │ │ State: applied │ │ Total Duration: 00:06:38 │ │ Total Cost: $0.107 │ │ Decisions Made: 12 │ │ Child Plans: 0 │ ╰──────────────────────────────────────────╯
╭─ Next Steps ──────╮ │ - Review git diff │ │ - Commit changes │ ╰───────────────────╯
✓ OK Changes applied
Example 3: Multi-File Refactoring with Invariants
Scenario: A senior engineer wants to refactor the authentication module to replace raw SQL queries with an ORM. The refactoring must maintain API backward compatibility and all existing tests must pass. Multiple invariants constrain the work.
Complexity: Intermediate-Advanced. Single project, multiple invariants, decision tree review.
Step 1: Define Invariants
$ agents invariant add --global "All public APIs must maintain backward compatibility"╭─ Invariant Added ──────────────────────────────────────────────────────╮ │ Invariant: All public APIs must maintain backward compatibility │ │ Scope: global │ │ ID: inv_01HXR3A1B │ ╰────────────────────────────────────────────────────────────────────────╯
✓ OK Invariant added
$ agents invariant add --project local/api-service
"Database queries must use the SQLAlchemy ORM, not raw SQL"╭─ Invariant Added ───────────────────────────────────────────────────────╮ │ Project: local/api-service │ │ Invariant: Database queries must use the SQLAlchemy ORM, not raw SQL │ │ Scope: project │ │ ID: inv_01HXR3A2C │ ╰─────────────────────────────────────────────────────────────────────────╯
✓ OK Invariant added
$ agents invariant list --project local/api-service --effective
╭─ Effective Invariants (Project local/api-service) ─────────────────────────────────────╮ │ ID Source Text │ │ ────────────── ──────── ────────────────────────────────────────────────────── │ │ inv_01HXR3A1B global All public APIs must maintain backward compatibility │ │ inv_01HXR3A2C project Database queries must use the SQLAlchemy ORM, not raw SQL │ ╰────────────────────────────────────────────────────────────────────────────────────────╯
✓ OK 2 effective invariants (1 global, 1 project)
Step 2: Create a Refactoring Action with Custom Actors
Create actors/refactoring-strategist.yaml:
name: local/refactoring-strategist
cleveragents:
version: "3.0"
actors:
strategist:
type: llm
config:
provider: anthropic
model: claude-3.5-sonnet
temperature: 0.2
system_prompt: |
You are an expert software architect specializing in database layer refactoring.
You analyze codebases to plan safe, incremental refactoring strategies.
You always preserve backward compatibility and plan for comprehensive testing.
Project: {{ context.project_name }}
Resources: {{ context.resources | join(', ') }}
memory_enabled: true
max_history: 100
$ agents actor add --config ./actors/refactoring-strategist.yaml╭─ Actor Added ────────────────────────────────────╮ │ Name: local/refactoring-strategist │ │ Provider: anthropic │ │ Model: claude-3.5-sonnet │ │ Default: yes │ │ Unsafe: no │ │ Type: graph │ ╰──────────────────────────────────────────────────╯
╭─ Config ────────────────────────────────────────────────────╮ │ Path: ./actors/refactoring-strategist.yaml │ │ Hash: 7a1e4c9 │ │ Options: 3 │ │ Nodes: 1 │ │ Edges: 0 │ ╰─────────────────────────────────────────────────────────────╯
╭─ Capabilities ────────────────╮ │ - database layer refactoring │ │ - incremental strategy │ │ - backward compatibility │ ╰───────────────────────────────╯
✓ OK Actor added
Create actions/refactor-to-orm.yaml:
name: local/refactor-to-orm
description: "Refactor raw SQL queries to use SQLAlchemy ORM"
long_description: |
Systematically replace all raw SQL queries in the target module with
SQLAlchemy ORM equivalents. Create ORM model definitions, replace
query construction, update transaction handling, and ensure all
existing behavior is preserved.
definition_of_done: |
- Zero raw SQL strings remain in the target module
- All ORM models are defined with proper relationships
- All existing tests pass without modification
- New integration tests verify ORM query correctness
- API response formats are unchanged
- Database migration script is generated
strategy_actor: local/refactoring-strategist
execution_actor: anthropic/claude-3.5-sonnet
estimation_actor: anthropic/claude-3.5-sonnet
reusable: true
state: available
automation_profile: cautious
args:
- name: target_module
type: string
required: true
description: "Module to refactor"
invariants:
- "Each file must be refactored in a separate commit-sized change"
- "ORM models must be defined before queries are converted"
- "All raw SQL must be replaced — no partial conversion"
$ agents action create --config ./actions/refactor-to-orm.yaml╭─ Action Created ─────────────────────────────────────────────╮ │ Name: local/refactor-to-orm │ │ ID: 01HXR3B1D2E3F4G5H6J7K8L9M0 │ │ State: available │ │ Strategy Actor: local/refactoring-strategist │ │ Execution Actor: anthropic/claude-3.5-sonnet │ │ Reusable: yes │ │ Read Only: no │ │ Created: 2026-02-11 09:14 │ ╰──────────────────────────────────────────────────────────────╯
╭─ Definition of Done ──────────────────────────────────────────╮ │ - Zero raw SQL strings remain in the target module │ │ - All ORM models are defined with proper relationships │ │ - All existing tests pass without modification │ │ - New integration tests verify ORM query correctness │ │ - API response formats are unchanged │ │ - Database migration script is generated │ ╰───────────────────────────────────────────────────────────────╯
╭─ Arguments ──────────────────────────────────────────────╮ │ Name Type Required Description │ │ ────────────── ────── ──────── ───────────────────── │ │ target_module string yes Module to refactor │ ╰──────────────────────────────────────────────────────────╯
╭─ Automation ──────────────────────────╮ │ Profile: cautious │ │ Source: action config │ ╰───────────────────────────────────────╯
╭─ Action Invariants ──────────────────────────────────────────────────────╮ │ 1. Each file must be refactored in a separate commit-sized change │ │ 2. ORM models must be defined before queries are converted │ │ 3. All raw SQL must be replaced — no partial conversion │ ╰──────────────────────────────────────────────────────────────────────────╯
╭─ Usage ───────────────────────────────────────────────────────────────────╮ │ agents plan use local/refactor-to-orm local/api-service │ │ --arg target_module="src/auth" │ ╰───────────────────────────────────────────────────────────────────────────╯
✓ OK Action created
Step 3: Execute with Cautious Profile
$ agents plan use \ --arg target_module="src/auth" \ local/refactor-to-orm local/api-service╭─ Plan Created ───────────────────────────────────────╮ │ Plan: 01HXR3C4D5E6F7G8H9J0K1L2M3 │ │ Action: local/refactor-to-orm │ │ Project: local/api-service │ │ Automation: cautious │ │ Attempt: 1 │ ╰──────────────────────────────────────────────────────╯
╭─ Inputs ────────────────────────╮ │ - target_module="src/auth" │ │ - automation_profile=cautious │ ╰─────────────────────────────────╯
╭─ Actors ───────────────────────────────────────╮ │ Strategy: local/refactoring-strategist │ │ Execution: anthropic/claude-3.5-sonnet │ │ Estimation: anthropic/claude-3.5-sonnet │ ╰────────────────────────────────────────────────╯
╭─ Automation ──────────────────────────────╮ │ Profile: cautious │ │ Source: action config │ │ Read-Only: no │ ╰───────────────────────────────────────────╯
╭─ Context ───────────────────────╮ │ Resources: 2 (repo, db) │ │ Indexed Files: 214 │ │ View: strategize │ │ Hot Token Budget: 16,000 │ ╰─────────────────────────────────╯
╭─ Next Steps ──────────────────────────────────────────╮ │ - agents plan execute 01HXR3C4D5E6F7G8H9J0K1L2M3 │ │ - agents plan status 01HXR3C4D5E6F7G8H9J0K1L2M3 │ │ - agents plan tree 01HXR3C4D5E6F7G8H9J0K1L2M3 │ ╰───────────────────────────────────────────────────────╯
✓ OK Plan created
Because the cautious profile uses intermediate confidence thresholds, the system:
- Enters Strategize automatically (threshold 0.7, confidence is high for well-understood tasks)
- Makes strategy decisions automatically when confidence >= 0.6, but pauses for complex architectural choices
- The Invariant Reconciliation Actor reconciles global ("backward compatibility"), project ("use ORM"), and action ("separate commits") invariants
# The system pauses at a decision where confidence is below threshold. # Review the decision tree:$ agents plan tree 01HXR3C4D5E6F7G8H9J0K1L2M3
╭─ Decision Tree ────────────────────────────────────────────────────────────────────────────╮ │ Plan: 01HXR3C4D5E6F7G8H9J0K1L2M3 │ │ ├─ [prompt_definition] "Refactor raw SQL in src/auth to SQLAlchemy ORM" │ │ ├─ [invariant_enforced] "All public APIs must maintain backward compatibility" │ │ ├─ [invariant_enforced] "Database queries must use the SQLAlchemy ORM" │ │ ├─ [invariant_enforced] "Each file refactored in a separate commit-sized change" │ │ ├─ [invariant_enforced] "ORM models must be defined before queries are converted" │ │ ├─ [strategy_choice] "Define ORM models first, then convert queries" (conf: 0.85) │ │ ├─ [strategy_choice] "Place models inline in route files" (conf: 0.48) ⏸ PAUSED │ │ └─ • (awaiting input — confidence below threshold) │ ╰────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Tree Summary ─────────────╮ │ Nodes: 8 │ │ Depth: 2 │ │ Child Plans: 0 │ │ Invariants: 4 │ │ Superseded: 0 (hidden) │ ╰────────────────────────────╯
╭─ Decision IDs (for correction) ──────────────────────╮ │ Root: 01HXR3D1A2B3C4D5E6F7G8H9 │ │ Invariant 1: 01HXR3D1B3C4D5E6F7G8H9J0 │ │ Invariant 2: 01HXR3D1C4D5E6F7G8H9J0K1 │ │ Invariant 3: 01HXR3D1D5E6F7G8H9J0K1L2 │ │ Invariant 4: 01HXR3D1E6F7G8H9J0K1L2M3 │ │ Strategy 1: 01HXR3D2E3F4G5H6I7J8K9L0 │ │ Strategy 2: 01HXR3D5F6G7H8I9J0K1L2M3 (paused) │ ╰──────────────────────────────────────────────────────╯
✓ OK Decision tree rendered
# The paused decision is about model placement. Inspect it:
$ agents plan explain 01HXR3D5F6G7H8I9J0K1L2M3
╭─ Decision ──────────────────────────────────────────────╮ │ ID: 01HXR3D5F6G7H8I9J0K1L2M3 │ │ Type: strategy_choice │ │ Question: Where should ORM model classes be placed? │ │ Chosen: Place models inline in route files │ │ Confidence: 0.48 │ │ Plan: 01HXR3C4D5E6F7G8H9J0K1L2M3 │ │ Sequence: 6 of 7 │ │ Created: 2026-02-11 09:15 │ ╰─────────────────────────────────────────────────────────╯
╭─ Alternatives Considered ─────────────────────────────────────╮ │ 1. Place models inline in route files (chosen) │ │ 2. Separate models/ directory with one model per file │ │ 3. Single models.py file alongside routes │ ╰───────────────────────────────────────────────────────────────╯
╭─ Impact ──────────────────────╮ │ Downstream Decisions: 4 │ │ Downstream Child Plans: 0 │ │ Artifacts Produced: 0 │ │ Correction Impact: high │ ╰───────────────────────────────╯
╭─ Rationale ────────────────────────────────────────────────────────╮ │ Placing models inline minimizes import changes and keeps related │ │ code co-located. However, this conflicts with project conventions │ │ and may make models harder to reuse across modules. The low │ │ confidence reflects genuine uncertainty about the trade-off. │ ╰────────────────────────────────────────────────────────────────────╯
╭─ Correction ───────────────────────────────────────────────────────╮ │ agents plan correct 01HXR3D5F6G7H8I9J0K1L2M3 │ │ --mode revert --guidance "Use separate models/ directory..." │ ╰────────────────────────────────────────────────────────────────────╯
✓ OK Decision explained
# Provide guidance via the plan prompt command
$ agents plan prompt 01HXR3C4D5E6F7G8H9J0K1L2M3
"Convert the User model first, then Session, then Token.
Use Alembic for the migration script."╭─ Guidance Added ─────────────────────────────────────────────────────────╮ │ Plan: 01HXR3C4D5E6F7G8H9J0K1L2M3 │ │ Guidance: Convert the User model first, then Session, then Token. │ │ Use Alembic for the migration script. │ │ Scope: next execution step │ │ Phase: strategize │ │ State: awaiting_input → processing │ ╰──────────────────────────────────────────────────────────────────────────╯
╭─ Decision Created ───────────────────────────╮ │ Type: user_intervention │ │ ID: 01HXR3E1A2B3C4D5E6F7G8H9 │ │ Parent: 01HXR3D5F6G7H8I9J0K1L2M3 │ ╰──────────────────────────────────────────────╯
╭─ Queue ────╮ │ Pending: 1 │ │ Applied: 0 │ ╰────────────╯
✓ OK Guidance queued
# Execution continues. Monitor progress:
$ agents plan status 01HXR3C4D5E6F7G8H9J0K1L2M3
╭─ Plan Status ───────────────────────────────────╮ │ Plan: 01HXR3C4D5E6F7G8H9J0K1L2M3 │ │ Phase: execute │ │ State: processing │ │ Action: local/refactor-to-orm │ │ Project: local/api-service │ │ Automation: cautious │ │ Attempt: 1 │ ╰─────────────────────────────────────────────────╯
╭─ Progress ───────╮ │ ✓ Strategize │ │ ⏳ Execute │ │ • Apply (queued) │ ╰──────────────────╯
╭─ Timing ──────────╮ │ Started: 09:14:22 │ │ Elapsed: 00:02:38 │ │ ETA: 00:05:10 │ ╰───────────────────╯
╭─ Execution Detail ──────────╮ │ Sandbox: git_worktree │ │ Tool Calls: 14 │ │ Files Modified: 4 │ │ Child Plans: 0/0 complete │ │ Checkpoints: 3 created │ ╰─────────────────────────────╯
╭─ Cost ───────────────╮ │ Tokens Used: 18,740 │ │ Cost So Far: $0.062 │ │ Estimated: $0.128 │ ╰──────────────────────╯
✓ OK Status refreshed
Step 4: Correct a Decision
During execution, you notice the strategy chose to inline the ORM models in the same file as the routes. You want them in a separate models/ directory:
# Find the decision that chose inline models$ agents plan tree --show-superseded 01HXR3C4D5E6F7G8H9J0K1L2M3
╭─ Decision Tree ────────────────────────────────────────────────────────────────────────────╮ │ Plan: 01HXR3C4D5E6F7G8H9J0K1L2M3 │ │ ├─ [prompt_definition] "Refactor raw SQL in src/auth to SQLAlchemy ORM" │ │ ├─ [invariant_enforced] "All public APIs must maintain backward compatibility" │ │ ├─ [invariant_enforced] "Database queries must use the SQLAlchemy ORM" │ │ ├─ [invariant_enforced] "Each file refactored in a separate commit-sized change" │ │ ├─ [invariant_enforced] "ORM models must be defined before queries are converted" │ │ ├─ [strategy_choice] "Define ORM models first, then convert queries" (conf: 0.85) │ │ ├─ [strategy_choice] "Place models inline in route files" (superseded) │ │ ├─ [user_intervention] "Convert User first, then Session, then Token" │ │ ├─ [strategy_choice] "Create User ORM model in routes/auth.py" (conf: 0.74) │ │ ├─ [tool_invocation] write_file src/auth/routes/auth.py │ │ └─ [tool_invocation] write_file src/auth/routes/session.py │ ╰────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Tree Summary ─────────────╮ │ Nodes: 11 │ │ Depth: 3 │ │ Child Plans: 0 │ │ Invariants: 4 │ │ Superseded: 1 (shown) │ ╰────────────────────────────╯
╭─ Decision IDs (for correction) ──────────────────────╮ │ Root: 01HXR3D1A2B3C4D5E6F7G8H9 │ │ Strategy 1: 01HXR3D2E3F4G5H6I7J8K9L0 │ │ Strategy 2: 01HXR3D5F6G7H8I9J0K1L2M3 (superseded) │ │ Intervention: 01HXR3E1A2B3C4D5E6F7G8H9 │ │ Strategy 3: 01HXR3F1A2B3C4D5E6F7G8H9 │ │ Tool 1: 01HXR3F2B3C4D5E6F7G8H9J0 │ │ Tool 2: 01HXR3F3C4D5E6F7G8H9J0K1 │ ╰──────────────────────────────────────────────────────╯
✓ OK Decision tree rendered
# Correct it — revert mode replays from that decision
$ agents plan correct 01HXR3F1A2B3C4D5E6F7G8H9
--mode revert
--guidance "Place ORM models in src/auth/models/ with one model per file.
Import them in src/auth/models/__init__.py." --yes╭─ Correction ────────────────────────────────────────────────╮ │ Mode: revert │ │ Impact: 2 decisions, 0 child plans, 2 artifacts │ │ New Decision: 01HXR3G1A2B3C4D5E6F7G8H9 │ │ Corrects: 01HXR3F1A2B3C4D5E6F7G8H9 │ │ Attempt: 2 │ │ Correction ID: 01HXR3CORR1A2B3C4D5E6 │ ╰─────────────────────────────────────────────────────────────╯
╭─ Affected Subtree ──────────────╮ │ Decisions Invalidated: 2 │ │ Child Plans Rolled Back: 0 │ │ Artifacts Archived: 2 │ │ Unaffected Decisions: 6 │ ╰─────────────────────────────────╯
╭─ Sandbox Rollback ─────────────╮ │ Checkpoint: cp_01HXR3C4 │ │ Files Reverted: 2 │ │ Status: restored │ ╰────────────────────────────────╯
╭─ Recompute ───────────────╮ │ Queued: re-execute │ │ ETA: 3m │ ╰───────────────────────────╯
╭─ History ───────────────────────────────────────────────╮ │ - Original decision superseded │ │ - Prior artifacts archived for comparison │ │ - agents plan diff --correction 01HXR3CORR1A2B3C4D5E6 │ ╰─────────────────────────────────────────────────────────╯
✓ OK Correction applied
# The affected subtree is recomputed. Review the new approach:
$ agents plan diff --correction 01HXR3CORR1A2B3C4D5E6
╭─ Correction Diff ────────────────────────────────────────────╮ │ Correction: 01HXR3CORR1A2B3C4D5E6 │ │ Original Decision: 01HXR3F1A2B3C4D5E6F7G8H9 │ │ Mode: revert │ │ Files Changed: 5 │ │ New Insertions: 84 │ │ New Deletions: 36 │ ╰──────────────────────────────────────────────────────────────╯
╭─ Comparison ──────────────────────────────────────────────────────────────╮ │ File Before (original) After (corrected) │ │ ────────────────────────────── ──────────────── ──────────────────── │ │ src/auth/routes/auth.py +42 -8 (inline) +6 -8 (imports only) │ │ src/auth/routes/session.py +28 -6 (inline) +4 -6 (imports only) │ │ src/auth/models/__init__.py (did not exist) (new file) │ │ src/auth/models/user.py (did not exist) (new file, +38) │ │ src/auth/models/session.py (did not exist) (new file, +26) │ ╰───────────────────────────────────────────────────────────────────────────╯
╭─ Patch Preview (corrected vs original) ───────────────────────────╮ │ --- a/src/auth/routes/auth.py (original) │ │ +++ b/src/auth/routes/auth.py (corrected) │ │ @@ -1,42 +1,6 @@ │ │ - class User(Base): # was inline │ │ - tablename = "users" │ │ - id = Column(Integer, primary_key=True) │ │ + from src.auth.models import User │ │ ... │ │ +++ b/src/auth/models/user.py (new file) │ │ @@ -0,0 +1,38 @@ │ │ + from sqlalchemy import Column, Integer, String │ │ + from sqlalchemy.orm import relationship │ │ + class User(Base): │ │ + tablename = "users" │ │ ... │ ╰───────────────────────────────────────────────────────────────────╯
✓ OK Correction diff generated
Step 5: Apply
$ agents plan diff 01HXR3C4D5E6F7G8H9J0K1L2M3╭─ Diff Summary ───────────────────────────────────────╮ │ Plan: 01HXR3C4D5E6F7G8H9J0K1L2M3 │ │ Project: local/api-service │ │ Files Changed: 9 │ │ Insertions: 186 │ │ Deletions: 94 │ │ Net Change: +92 lines │ ╰──────────────────────────────────────────────────────╯
╭─ Files ───────────────────────────────────────────────╮ │ Path Change Status │ │ ────────────────────────────── ──────── ────────── │ │ src/auth/models/__init__.py +8 new file │ │ src/auth/models/user.py +38 new file │ │ src/auth/models/session.py +26 new file │ │ src/auth/models/token.py +22 new file │ │ src/auth/routes/auth.py +12 -28 modified │ │ src/auth/routes/session.py +8 -22 modified │ │ src/auth/routes/token.py +6 -18 modified │ │ src/auth/db.py +14 -26 modified │ │ alembic/versions/001_orm.py +52 new file │ ╰───────────────────────────────────────────────────────╯
╭─ Risk Assessment ────────────────╮ │ API Compatibility: preserved │ │ Test Coverage: maintained │ │ Breaking Changes: none detected │ ╰──────────────────────────────────╯
✓ OK Diff generated
$ agents plan apply --yes 01HXR3C4D5E6F7G8H9J0K1L2M3
╭─ Apply Summary ───────────────────────────────────╮ │ Plan: 01HXR3C4D5E6F7G8H9J0K1L2M3 │ │ Artifacts: 9 files updated │ │ Changes: 186 insertions, 94 deletions │ │ Project: local/api-service │ │ Applied At: 2026-02-11 09:23 │ ╰───────────────────────────────────────────────────╯
╭─ Validation ───────────────────────╮ │ Tests: passed (47/47) │ │ Lint: passed (0 warnings) │ │ Type Check: passed (0 errors) │ │ Duration: 18.6s │ ╰────────────────────────────────────╯
╭─ Sandbox Cleanup ─────────╮ │ Worktree: removed │ │ Branch: merged to main │ │ Checkpoint: archived │ ╰───────────────────────────╯
╭─ Plan Lifecycle ───────────────────────────────╮ │ Phase: apply │ │ State: applied │ │ Total Duration: 00:08:42 │ │ Total Cost: $0.128 │ │ Decisions Made: 11 │ │ Corrections: 1 │ │ Child Plans: 0 │ ╰────────────────────────────────────────────────╯
╭─ Next Steps ──────╮ │ - Review git diff │ │ - Commit changes │ ╰───────────────────╯
✓ OK Changes applied
Example 4: Multi-Project Dependency Update
Scenario: An organization has three microservices that share a common library. The library has a breaking change that requires coordinated updates across all three services. The update must be validated in each service independently before any changes are applied.
Complexity: Advanced. Multiple projects, child plans, coordinated apply.
Step 1: Register All Projects and Resources
# Register the common-lib resource (full output shown)$ agents resource add git-checkout local/common-lib-repo --path /repos/common-lib --branch main
╭─ Resource ──────────────────────────────────────╮ │ Name: local/common-lib-repo │ │ ID: 01HXR4A1B2C3D4E5F6G7H8J9K0 │ │ Type: git-checkout │ │ Physical/Virtual: physical │ │ Path: /repos/common-lib │ │ Branch: main │ │ Created: 2026-02-11 10:00 │ ╰─────────────────────────────────────────────────╯
╭─ Auto-discovered Children ─────────────────────────────────────────╮ │ ID Type Status │ │ ─────────────── ────────────── ───────────────── │ │ 01HXR4A1B2C3… git created │ │ 01HXR4A1B2C4… git-remote created │ │ 01HXR4A1B2C5… git-branch created │ │ + 31 git-commit resources │ │ + 186 git-tree-entry resources │ │ + 2 fs-directory + 14 fs-file │ ╰────────────────────────────────────────────────────────────────────╯
╭─ Capabilities ─────────────────╮ │ Readable: yes │ │ Writable: yes │ │ Sandboxable: yes │ │ Checkpointable: yes │ │ Sandbox Strategy: git_worktree │ ╰────────────────────────────────╯
✓ OK Resource registered (236 child resources discovered)
# Register the remaining 3 service resources (condensed output)
$ agents resource add git-checkout local/user-service-repo --path /repos/user-service --branch main ✓ OK Resource registered — Name: local/user-service-repo (312 children)
$ agents resource add git-checkout local/order-service-repo --path /repos/order-service --branch main ✓ OK Resource registered — Name: local/order-service-repo (287 children)
$ agents resource add git-checkout local/billing-service-repo --path /repos/billing-service --branch main ✓ OK Resource registered — Name: local/billing-service-repo (264 children)
# Create the common-lib project (full output shown)$ agents project create -d "Shared common library"
--resource local/common-lib-repo local/common-lib╭─ Project Created ─────────────────────────╮ │ Name: local/common-lib │ │ Description: Shared common library │ │ Type: local │ │ Created: 2026-02-11 10:01 │ ╰───────────────────────────────────────────╯
╭─ Linked Resources ───────────────────────────────────────╮ │ Resource Type Read-Only │ │ ───────────────────── ────────────── ───────── │ │ local/common-lib-repo git-checkout no │ ╰──────────────────────────────────────────────────────────╯
╭─ Defaults ──────────────────────────────╮ │ Sandbox: git_worktree │ │ Validations: 0 │ │ Context Filters: none │ │ Automation Profile: (inherits global) │ ╰─────────────────────────────────────────╯
✓ OK Project created
# Create the remaining 3 service projects (condensed output)
$ agents project create -d "User microservice"
--resource local/user-service-repo local/user-service ✓ OK Project created — Name: local/user-service$ agents project create -d "Order microservice"
--resource local/order-service-repo local/order-service ✓ OK Project created — Name: local/order-service
$ agents project create -d "Billing microservice"
--resource local/billing-service-repo local/billing-service ✓ OK Project created — Name: local/billing-service
# Register a shared validation — first one with full output$ agents validation add --config validations/pytest-mypy.yaml --required local/pytest-mypy
╭─ Validation Registered ─────────────────────────────╮ │ Name: local/pytest-mypy │ │ Command: pytest tests/ && mypy src/ │ │ Mode: required │ │ Timeout: 300s │ ╰─────────────────────────────────────────────────────╯
✓ OK Validation registered
# Attach the validation to all 4 projects
$ agents validation attach --project local/common-lib local/pytest-mypy ✓ OK Validation attached to project local/common-lib
$ agents validation attach --project local/user-service local/pytest-mypy ✓ OK Validation attached to project local/user-service
$ agents validation attach --project local/order-service local/pytest-mypy ✓ OK Validation attached to project local/order-service
$ agents validation attach --project local/billing-service local/pytest-mypy ✓ OK Validation attached to project local/billing-service
Step 2: Create a Coordinated Update Action
Create actions/coordinated-dependency-update.yaml:
name: local/coordinated-dep-update
description: "Update a shared dependency across multiple projects"
long_description: |
When a shared library introduces a breaking change, this action coordinates
the update across all dependent projects. It analyzes the breaking changes,
creates per-project update plans as child plans, validates each independently,
and produces a coordinated changeset.
definition_of_done: |
- The shared library is updated to the target version
- All dependent projects compile and pass tests with the new version
- API contracts between services remain compatible
- Migration guides are generated for each project if needed
- All changes can be applied atomically or rolled back as a unit
strategy_actor: anthropic/claude-3.5-sonnet
execution_actor: anthropic/claude-3.5-sonnet
automation_profile: supervised
reusable: true
state: available
args:
- name: library_name
type: string
required: true
description: "Name of the shared library to update"
- name: target_version
type: string
required: true
description: "Target version to update to"
- name: changelog_url
type: string
required: false
description: "URL to the library changelog or migration guide"
invariants:
- "Each dependent project must be updated in its own child plan"
- "All child plans must pass validation before any can be applied"
- "The library update in common-lib must be applied first"
$ agents action create --config ./actions/coordinated-dependency-update.yaml╭─ Action Created ─────────────────────────────────────────────╮ │ Name: local/coordinated-dep-update │ │ ID: 01HXR4B1D2E3F4G5H6J7K8L9M0 │ │ State: available │ │ Strategy Actor: anthropic/claude-3.5-sonnet │ │ Execution Actor: anthropic/claude-3.5-sonnet │ │ Reusable: yes │ │ Read Only: no │ │ Created: 2026-02-11 10:02 │ ╰──────────────────────────────────────────────────────────────╯
╭─ Definition of Done ──────────────────────────────────────────────────╮ │ - The shared library is updated to the target version │ │ - All dependent projects compile and pass tests with the new version │ │ - API contracts between services remain compatible │ │ - Migration guides are generated for each project if needed │ │ - All changes can be applied atomically or rolled back as a unit │ ╰───────────────────────────────────────────────────────────────────────╯
╭─ Arguments ──────────────────────────────────────────────────────────────────────╮ │ Name Type Required Description │ │ ────────────── ────── ──────── ────────────────────────────────────────── │ │ library_name string yes Name of the shared library to update │ │ target_version string yes Target version to update to │ │ changelog_url string no URL to the library changelog or migration │ ╰──────────────────────────────────────────────────────────────────────────────────╯
╭─ Automation ──────────────────────────╮ │ Profile: supervised │ │ Source: action config │ ╰───────────────────────────────────────╯
╭─ Action Invariants ──────────────────────────────────────────────────────────────╮ │ 1. Each dependent project must be updated in its own child plan │ │ 2. All child plans must pass validation before any can be applied │ │ 3. The library update in common-lib must be applied first │ ╰──────────────────────────────────────────────────────────────────────────────────╯
╭─ Usage ───────────────────────────────────────────────────────────────────────────╮ │ agents plan use local/coordinated-dep-update local/common-lib │ │ --arg library_name="authlib" --arg target_version="2.0.0" │ ╰───────────────────────────────────────────────────────────────────────────────────╯
✓ OK Action created
Step 3: Execute Across Multiple Projects
# Create a plan targeting all four projects$ agents plan use
--automation-profile supervised
--arg library_name="authlib"
--arg target_version="2.0.0"
--arg changelog_url="https://authlib.org/changelog/2.0"
local/coordinated-dep-update
local/common-lib local/user-service local/order-service local/billing-service╭─ Plan Created ────────────────────────────────────────────╮ │ Plan: 01HXR4D5E6F7G8H9J0K1L2M3N4 │ │ Action: local/coordinated-dep-update │ │ Projects: local/common-lib, local/user-service, │ │ local/order-service, local/billing-service │ │ Automation: supervised │ │ Attempt: 1 │ ╰───────────────────────────────────────────────────────────╯
╭─ Inputs ───────────────────────────────────────────────────────────╮ │ - library_name="authlib" │ │ - target_version="2.0.0" │ │ - changelog_url="https://authlib.org/changelog/2.0" │ │ - automation_profile=supervised │ ╰────────────────────────────────────────────────────────────────────╯
╭─ Actors ───────────────────────────────────────╮ │ Strategy: anthropic/claude-3.5-sonnet │ │ Execution: anthropic/claude-3.5-sonnet │ ╰────────────────────────────────────────────────╯
╭─ Automation ──────────────────────────────╮ │ Profile: supervised │ │ Source: CLI override │ │ Read-Only: no │ ╰───────────────────────────────────────────╯
╭─ Context ──────────────────────────╮ │ Resources: 4 (git-checkout) │ │ Indexed Files: 1,099 │ │ View: strategize │ │ Hot Token Budget: 24,000 │ ╰────────────────────────────────────╯
╭─ Next Steps ──────────────────────────────────────────────╮ │ - agents plan tree 01HXR4D5E6F7G8H9J0K1L2M3N4 │ │ - agents plan status 01HXR4D5E6F7G8H9J0K1L2M3N4 │ │ - agents plan execute 01HXR4D5E6F7G8H9J0K1L2M3N4 │ ╰───────────────────────────────────────────────────────────╯
✓ OK Plan created
The system creates a parent plan. Strategize runs automatically (supervised profile). It produces a strategy that:
- Analyzes the breaking changes in authlib 2.0.0
- Plans child plans for each project
- Orders the updates: common-lib first, then services in parallel
# Review the strategy before execution begins$ agents plan tree 01HXR4D5E6F7G8H9J0K1L2M3N4
╭─ Decision Tree ─────────────────────────────────────────────────────────────────────────────────────╮ │ Plan: 01HXR4D5E6F7G8H9J0K1L2M3N4 │ │ ├─ [prompt_definition] "Update authlib to 2.0.0 across 4 projects" │ │ ├─ [invariant_enforced] "Each dependent project must be updated in its own child plan" │ │ ├─ [invariant_enforced] "All child plans must pass validation before any can be applied" │ │ ├─ [invariant_enforced] "The library update in common-lib must be applied first" │ │ ├─ [strategy_choice] "Analyze changelog for breaking API changes" (conf: 0.92) │ │ ├─ [strategy_choice] "Update common-lib first, then services in parallel" (conf: 0.88) │ │ ├─ [strategy_choice] "Create 4 child plans: common-lib, user-svc, order-svc, billing-svc" │ │ │ (conf: 0.90) │ │ └─ • (awaiting execution approval — supervised profile) │ ╰─────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Tree Summary ──────────────╮ │ Nodes: 8 │ │ Depth: 2 │ │ Child Plans: 0 (planned: 4) │ │ Invariants: 3 │ │ Superseded: 0 (hidden) │ ╰─────────────────────────────╯
╭─ Decision IDs (for correction) ──────────────────────────╮ │ Root: 01HXR4D6A1B2C3D4E5F6G7H8 │ │ Invariant 1: 01HXR4D6B2C3D4E5F6G7H8J9 │ │ Invariant 2: 01HXR4D6C3D4E5F6G7H8J9K0 │ │ Invariant 3: 01HXR4D6D4E5F6G7H8J9K0L1 │ │ Strategy 1: 01HXR4D7E5F6G7H8I9J0K1L2 │ │ Strategy 2: 01HXR4D7F6G7H8I9J0K1L2M3 │ │ Strategy 3: 01HXR4D7G7H8I9J0K1L2M3N4 │ ╰──────────────────────────────────────────────────────────╯
✓ OK Decision tree rendered
$ agents plan status 01HXR4D5E6F7G8H9J0K1L2M3N4
╭─ Plan Status ──────────────────────────────────────────╮ │ Plan: 01HXR4D5E6F7G8H9J0K1L2M3N4 │ │ Phase: strategize → execute (pending approval) │ │ State: awaiting_input │ │ Action: local/coordinated-dep-update │ │ Projects: 4 │ │ Automation: supervised │ │ Attempt: 1 │ ╰────────────────────────────────────────────────────────╯
╭─ Progress ────────────────────────────╮ │ ✓ Strategize │ │ ⏸ Execute (awaiting approval) │ │ • Apply (queued) │ ╰───────────────────────────────────────╯
╭─ Timing ──────────╮ │ Started: 10:02:14 │ │ Elapsed: 00:01:42 │ │ ETA: 00:12:00 │ ╰───────────────────╯
╭─ Cost ───────────────╮ │ Tokens Used: 8,420 │ │ Cost So Far: $0.028 │ │ Estimated: $0.340 │ ╰──────────────────────╯
✓ OK Status refreshed
# Manually approve execution (supervised requires explicit approval)
$ agents plan execute 01HXR4D5E6F7G8H9J0K1L2M3N4
╭─ Execution Started ──────────────────────────────────────╮ │ Plan: 01HXR4D5E6F7G8H9J0K1L2M3N4 │ │ Phase: execute │ │ State: processing │ │ Automation: supervised │ ╰──────────────────────────────────────────────────────────╯
╭─ Child Plans Spawned ───────────────────────────────────────────────────╮ │ Plan ID Project Status │ │ ────────────────────────── ─────────────────── ──────────────── │ │ 01HXR4C1D2E3F4G5H6I7J8K9 local/common-lib executing │ │ 01HXR4C2E3F4G5H6I7J8K9L0 local/user-service waiting (common) │ │ 01HXR4C3F4G5H6I7J8K9L0M1 local/order-service waiting (common) │ │ 01HXR4C4G5H6I7J8K9L0M1N2 local/billing-service waiting (common) │ ╰─────────────────────────────────────────────────────────────────────────╯
╭─ Execution Order ──────────────────────────────────────────╮ │ Phase 1: common-lib (blocking) │ │ Phase 2: user-service, order-service, billing-service │ │ (parallel, after common-lib completes) │ ╰────────────────────────────────────────────────────────────╯
✓ OK Execution started (4 child plans spawned)
Step 4: Review and Apply in Order
Once all child plans complete execution and pass validation, review and apply them in dependency order.
# Review common-lib changes first (must be applied before services)$ agents plan diff 01HXR4C1D2E3F4G5H6I7J8K9
╭─ Diff Summary ───────────────────────────────────────╮ │ Plan: 01HXR4C1D2E3F4G5H6I7J8K9 │ │ Project: local/common-lib │ │ Files Changed: 4 │ │ Insertions: 62 │ │ Deletions: 38 │ │ Net Change: +24 lines │ ╰──────────────────────────────────────────────────────╯
╭─ Files ───────────────────────────────────────────────────────╮ │ Path Change Status │ │ ──────────────────────────────── ──────── ────────── │ │ requirements.txt +1 -1 modified │ │ src/common/auth.py +28 -19 modified │ │ src/common/token_validator.py +18 -12 modified │ │ tests/test_auth.py +15 -6 modified │ ╰───────────────────────────────────────────────────────────────╯
╭─ Patch Preview ───────────────────────────────────────────────────╮ │ --- a/requirements.txt │ │ +++ b/requirements.txt │ │ @@ -3,1 +3,1 @@ │ │ - authlib==1.3.2 │ │ + authlib==2.0.0 │ │ --- a/src/common/auth.py │ │ +++ b/src/common/auth.py │ │ @@ -1,6 +1,8 @@ │ │ - from authlib.integrations.requests_client import OAuth2Session │ │ + from authlib.integrations.httpx_client import AsyncOAuth2Client │ │ ... │ ╰───────────────────────────────────────────────────────────────────╯
╭─ Validation ────────────────────────╮ │ pytest tests/ && mypy src/: passed │ │ Duration: 12.4s │ ╰─────────────────────────────────────╯
✓ OK Diff generated
$ agents plan apply --yes 01HXR4C1D2E3F4G5H6I7J8K9
╭─ Apply Summary ──────────────────────────────────────╮ │ Plan: 01HXR4C1D2E3F4G5H6I7J8K9 │ │ Artifacts: 4 files updated │ │ Changes: 62 insertions, 38 deletions │ │ Project: local/common-lib │ │ Applied At: 2026-02-11 10:14 │ ╰──────────────────────────────────────────────────────╯
╭─ Validation ───────────────────────╮ │ Tests: passed (24/24) │ │ Type Check: passed (0 errors) │ │ Duration: 12.4s │ ╰────────────────────────────────────╯
╭─ Sandbox Cleanup ─────────╮ │ Worktree: removed │ │ Branch: merged to main │ │ Checkpoint: archived │ ╰───────────────────────────╯
╭─ Plan Lifecycle ───────────────────────────────╮ │ Phase: apply │ │ State: applied │ │ Total Duration: 00:03:18 │ │ Total Cost: $0.064 │ ╰────────────────────────────────────────────────╯
✓ OK Changes applied
# Once common-lib is applied, review the 3 service child plans. # Full diff shown for user-service; order-service and billing-service condensed.$ agents plan diff 01HXR4C2E3F4G5H6I7J8K9L0
╭─ Diff Summary ───────────────────────────────────────╮ │ Plan: 01HXR4C2E3F4G5H6I7J8K9L0 │ │ Project: local/user-service │ │ Files Changed: 6 │ │ Insertions: 84 │ │ Deletions: 52 │ │ Net Change: +32 lines │ ╰──────────────────────────────────────────────────────╯
╭─ Files ───────────────────────────────────────────────────────╮ │ Path Change Status │ │ ──────────────────────────────── ──────── ────────── │ │ requirements.txt +1 -1 modified │ │ src/auth/middleware.py +22 -14 modified │ │ src/auth/oauth.py +18 -12 modified │ │ src/auth/token.py +16 -11 modified │ │ tests/test_middleware.py +14 -8 modified │ │ tests/test_oauth.py +13 -6 modified │ ╰───────────────────────────────────────────────────────────────╯
╭─ Validation ────────────────────────╮ │ pytest tests/ && mypy src/: passed │ │ Duration: 18.2s │ ╰─────────────────────────────────────╯
✓ OK Diff generated
$ agents plan diff 01HXR4C3F4G5H6I7J8K9L0M1
╭─ Diff Summary ───────────────────────────────────────╮ │ Plan: 01HXR4C3F4G5H6I7J8K9L0M1 │ │ Project: local/order-service │ │ Files Changed: 5 │ │ Insertions: 71 │ │ Deletions: 46 │ │ Net Change: +25 lines │ ╰──────────────────────────────────────────────────────╯
╭─ Validation ────────────────────────╮ │ pytest tests/ && mypy src/: passed │ │ Duration: 14.7s │ ╰─────────────────────────────────────╯
✓ OK Diff generated
$ agents plan diff 01HXR4C4G5H6I7J8K9L0M1N2
╭─ Diff Summary ───────────────────────────────────────╮ │ Plan: 01HXR4C4G5H6I7J8K9L0M1N2 │ │ Project: local/billing-service │ │ Files Changed: 4 │ │ Insertions: 58 │ │ Deletions: 34 │ │ Net Change: +24 lines │ ╰──────────────────────────────────────────────────────╯
╭─ Validation ────────────────────────╮ │ pytest tests/ && mypy src/: passed │ │ Duration: 11.3s │ ╰─────────────────────────────────────╯
✓ OK Diff generated
# Apply all 3 service child plans
$ agents plan apply --yes 01HXR4C2E3F4G5H6I7J8K9L0
╭─ Apply Summary ──────────────────────────────────────╮ │ Plan: 01HXR4C2E3F4G5H6I7J8K9L0 │ │ Artifacts: 6 files updated │ │ Changes: 84 insertions, 52 deletions │ │ Project: local/user-service │ │ Applied At: 2026-02-11 10:16 │ ╰──────────────────────────────────────────────────────╯
╭─ Validation ───────────────────────╮ │ Tests: passed (89/89) │ │ Type Check: passed (0 errors) │ │ Duration: 18.2s │ ╰────────────────────────────────────╯
╭─ Sandbox Cleanup ─────────╮ │ Worktree: removed │ │ Branch: merged to main │ │ Checkpoint: archived │ ╰───────────────────────────╯
╭─ Plan Lifecycle ───────────────────────────────╮ │ Phase: apply │ │ State: applied │ │ Total Duration: 00:04:12 │ │ Total Cost: $0.082 │ ╰────────────────────────────────────────────────╯
✓ OK Changes applied
$ agents plan apply --yes 01HXR4C3F4G5H6I7J8K9L0M1
╭─ Apply Summary ──────────────────────────────────────╮ │ Plan: 01HXR4C3F4G5H6I7J8K9L0M1 │ │ Artifacts: 5 files updated │ │ Changes: 71 insertions, 46 deletions │ │ Project: local/order-service │ │ Applied At: 2026-02-11 10:17 │ ╰──────────────────────────────────────────────────────╯
╭─ Validation ───────────────────────╮ │ Tests: passed (76/76) │ │ Type Check: passed (0 errors) │ │ Duration: 14.7s │ ╰────────────────────────────────────╯
╭─ Plan Lifecycle ───────────────────────────────╮ │ Phase: apply │ │ State: applied │ │ Total Duration: 00:03:48 │ │ Total Cost: $0.074 │ ╰────────────────────────────────────────────────╯
✓ OK Changes applied
$ agents plan apply --yes 01HXR4C4G5H6I7J8K9L0M1N2
╭─ Apply Summary ──────────────────────────────────────╮ │ Plan: 01HXR4C4G5H6I7J8K9L0M1N2 │ │ Artifacts: 4 files updated │ │ Changes: 58 insertions, 34 deletions │ │ Project: local/billing-service │ │ Applied At: 2026-02-11 10:18 │ ╰──────────────────────────────────────────────────────╯
╭─ Validation ───────────────────────╮ │ Tests: passed (63/63) │ │ Type Check: passed (0 errors) │ │ Duration: 11.3s │ ╰────────────────────────────────────╯
╭─ Plan Lifecycle ───────────────────────────────╮ │ Phase: apply │ │ State: applied │ │ Total Duration: 00:03:22 │ │ Total Cost: $0.068 │ ╰────────────────────────────────────────────────╯
╭─ Parent Plan Complete ──────────────────────────────────╮ │ Plan: 01HXR4D5E6F7G8H9J0K1L2M3N4 │ │ Phase: apply │ │ State: applied │ │ Child Plans: 4/4 complete │ │ Total Duration: 00:16:04 │ │ Total Cost: $0.340 │ │ All Validations: passed │ ╰─────────────────────────────────────────────────────────╯
✓ OK Changes applied
Example 5: Database Schema Migration with Safety Nets
Scenario: A team needs to add a last_login_at column to a users table with 10 million rows, backfill it from an audit log, and update the application code — all without downtime.
Complexity: Advanced. Mixed resource types (git + database), custom tools, checkpointing critical.
Step 1: Register the Database as a Resource
Register the custom resource type, then create a resource instance and link it to the project:
# Register the custom postgres-db resource type $ agents resource type add --config ./resource-types/postgres-db.yaml╭─ Resource Type ─────────────────────────╮ │ Name: local/postgres-db │ │ Physical/Virtual: physical │ │ User Addable: yes │ │ Registered: 2026-02-11 09:01 │ ╰─────────────────────────────────────────╯
╭─ CLI Arguments ──────────────────────────────────────────╮ │ Argument Required Description │ │ ──────────── ──────── ────────────────────────────── │ │ --host yes Database hostname │ │ --port no Port (default: 5432) │ │ --database yes Database name │ │ --schema no Schema (default: public) │ ╰──────────────────────────────────────────────────────────╯
╭─ Sandbox ──────────────────────────────╮ │ Strategy: transaction_rollback │ │ Handler: PostgresHandler (custom) │ ╰────────────────────────────────────────╯
✓ OK Resource type registered ℹ New subcommand available: agents resource add local/postgres-db
# Register the production database instance $ agents resource add local/postgres-db local/prod-users-db \ --host db.internal.example.com \ --port 5432 \ --database users_db \ --schema public╭─ Resource ───────────────────────────────────────╮ │ Name: local/prod-users-db │ │ ID: 01HXR5D3E4F5G6H7J8K9L0M1N2 │ │ Type: local/postgres-db │ │ Physical/Virtual: physical │ │ Host: db.internal.example.com │ │ Port: 5432 │ │ Database: users_db │ │ Schema: public │ │ Created: 2026-02-11 09:02 │ ╰──────────────────────────────────────────────────╯
╭─ Capabilities ──────────────────────────╮ │ Readable: yes │ │ Writable: yes │ │ Sandboxable: yes │ │ Checkpointable: yes │ │ Sandbox Strategy: transaction_rollback │ ╰─────────────────────────────────────────╯
✓ OK Resource registered
# Link the database resource to the project $ agents project link-resource local/api-service local/prod-users-db╭─ Resource Linked ────────────────────────────╮ │ Project: local/api-service │ │ Resource: local/prod-users-db │ │ Type: local/postgres-db │ │ Read-Only: no │ ╰──────────────────────────────────────────────╯
╭─ Access ─────────────────╮ │ Read: allowed │ │ Write: allowed │ │ Apply: requires approval │ ╰──────────────────────────╯
╭─ Project Resources ─────────────────────────────────────────────╮ │ ID Type Status │ │ ────────────────────────────── ──────────────── ────────── │ │ 01HXR1A1B2C3D4E5F6G7H8J9K0 git-checkout linked │ │ 01HXR5D3E4F5G6H7J8K9L0M1N2 local/postgres-db linked (new) │ ╰─────────────────────────────────────────────────────────────────╯
✓ OK Resource linked to project
Step 2: Create a Custom Skill with Database Tools
Create skills/database-ops.yaml:
name: local/database-ops
description: "Safe database operations with transaction support"
tools:
- name: query_db
description: "Execute a read-only SQL query and return results"
input_schema:
type: object
properties:
sql:
type: string
description: "SQL SELECT query"
params:
type: array
description: "Query parameters"
required: [sql]
writes: false
checkpointable: false
resource_slots:
- name: database
type: local/postgres-db
binding: contextual
-
name: execute_migration description: "Execute a DDL migration within a transaction" input_schema: type: object properties: up_sql: type: string description: "Forward migration SQL" down_sql: type: string description: "Rollback migration SQL" description: type: string description: "Human-readable description" required: [up_sql, down_sql, description] writes: true checkpointable: true resource_slots:
- name: database type: local/postgres-db binding: contextual
name: backfill_column description: "Batch-update a column using a source query" input_schema: type: object properties: table: type: string column: type: string source_query: type: string description: "Query that returns (id, value) pairs" batch_size: type: integer default: 10000 required: [table, column, source_query] writes: true checkpointable: true resource_slots:
name: database type: local/postgres-db binding: contextual
$ agents skill add --config ./skills/database-ops.yaml╭─ Skill Registered ─────────────────────────────────────────────╮ │ Name: local/database-ops │ │ Description: Safe database operations with transaction support │ │ Config: ./skills/database-ops.yaml │ │ Created: 2026-02-11 09:04 │ ╰────────────────────────────────────────────────────────────────╯
╭─ Tool Sources ──────────────────────────────────────╮ │ Source Count Details │ │ ──────── ───── ──────────────────────────── │ │ custom 3 query_db, execute_migration, │ │ backfill_column │ │ ──────── ───── ──────────────────────────── │ │ Total: 3 │ ╰─────────────────────────────────────────────────────╯
╭─ Resource Slots ───────────────────────────────────╮ │ Tool Slot Type │ │ ───────────────── ──────── ────────────────── │ │ query_db database local/postgres-db │ │ execute_migration database local/postgres-db │ │ backfill_column database local/postgres-db │ ╰────────────────────────────────────────────────────╯
✓ OK Skill registered with 3 tools
Step 3: Create the Migration Action and Execute the Plan
Create actions/add-column-with-backfill.yaml:
name: local/add-column-with-backfill
description: "Add a column, backfill data, and update application code"
definition_of_done: |
- Database migration adds the column with a sensible default
- Backfill completes for all rows using the specified source
- Application code reads/writes the new column
- All tests pass with the new schema
- A rollback migration is available and tested
strategy_actor: anthropic/claude-3.5-sonnet
execution_actor: anthropic/claude-3.5-sonnet
automation_profile: review
reusable: true
state: available
args:
- name: table_name
type: string
required: true
- name: column_name
type: string
required: true
- name: column_type
type: string
required: true
- name: backfill_source
type: string
required: true
description: "Description of where to source backfill data"
invariants:
- "Migration must be backward-compatible (add column, don't rename or drop)"
- "Backfill must be batched to avoid locking the table"
- "Rollback migration must be provided and tested"
- "Application code must handle both old (null) and new values gracefully"
$ agents action create --config ./actions/add-column-with-backfill.yaml╭─ Action Created ────────────────────────────────────╮ │ Name: local/add-column-with-backfill │ │ State: available │ │ Strategy Actor: anthropic/claude-3.5-sonnet │ │ Execution Actor: anthropic/claude-3.5-sonnet │ │ Reusable: yes │ │ Read Only: no │ │ Config: ./actions/add-column-with-backfill.yaml │ │ Created: 2026-02-11 09:05 │ ╰─────────────────────────────────────────────────────╯
╭─ Definition of Done ─────────────────────────────────────────────╮ │ - Database migration adds the column with a sensible default │ │ - Backfill completes for all rows using the specified source │ │ - Application code reads/writes the new column │ │ - All tests pass with the new schema │ │ - A rollback migration is available and tested │ ╰──────────────────────────────────────────────────────────────────╯
╭─ Arguments ────────────────────────────────────────────────────────────╮ │ Name Type Required Description │ │ ──────────────── ────── ──────── ───────────────────────────────── │ │ table_name string yes (none) │ │ column_name string yes (none) │ │ column_type string yes (none) │ │ backfill_source string yes Where to source backfill data │ ╰────────────────────────────────────────────────────────────────────────╯
╭─ Invariants ──────────────────────────────────────────────────────────────────────╮ │ 1. Migration must be backward-compatible (add column, don't rename or drop) │ │ 2. Backfill must be batched to avoid locking the table │ │ 3. Rollback migration must be provided and tested │ │ 4. Application code must handle both old (null) and new values gracefully │ ╰───────────────────────────────────────────────────────────────────────────────────╯
╭─ Automation ──────────────────────╮ │ Profile: review │ │ Source: action config │ ╰───────────────────────────────────╯
╭─ Usage ─────────────────────────────────────────────────────────────────────────╮ │ agents plan use local/add-column-with-backfill local/api-service │ │ --arg table_name=TABLE --arg column_name=COL │ │ --arg column_type=TYPE --arg backfill_source=SOURCE │ ╰─────────────────────────────────────────────────────────────────────────────────╯
✓ OK Action created
$ agents plan use local/add-column-with-backfill local/api-service \ --arg table_name="users" \ --arg column_name="last_login_at" \ --arg column_type="TIMESTAMP WITH TIME ZONE" \ --arg backfill_source="audit_log table, event_type='login', max(created_at) per user_id"╭─ Plan Created ──────────────────────────────────────╮ │ Plan ID: 01HXR5E6F7G8H9J0K1L2M3N4O5 │ │ Phase: strategize │ │ Action: local/add-column-with-backfill │ │ Project: local/api-service │ │ Automation: review │ │ Attempt: 1 │ ╰─────────────────────────────────────────────────────╯
╭─ Inputs ────────────────────────────────────────────────────────────────────────────╮ │ - table_name=users │ │ - column_name=last_login_at │ │ - column_type=TIMESTAMP WITH TIME ZONE │ │ - backfill_source=audit_log table, event_type='login', max(created_at) per user_id │ ╰─────────────────────────────────────────────────────────────────────────────────────╯
╭─ Actors ───────────────────────────────────╮ │ Strategy: anthropic/claude-3.5-sonnet │ │ Execution: anthropic/claude-3.5-sonnet │ │ Estimation: (none) │ ╰────────────────────────────────────────────╯
╭─ Plan Invariants ──────────────────────────────────────────────────────────────╮ │ Scope Source Invariant │ │ ──────── ─────── ────────────────────────────────────────────────────── │ │ plan action Migration must be backward-compatible │ │ plan action Backfill must be batched to avoid locking the table │ │ plan action Rollback migration must be provided and tested │ │ plan action Application code must handle both old (null) and new values │ ╰────────────────────────────────────────────────────────────────────────────────╯
╭─ Context ──────────────────────────────╮ │ Resources: 2 (git-checkout, postgres) │ │ Indexed Files: 347 │ │ View: strategize │ │ Hot Token Budget: 12,000 │ ╰────────────────────────────────────────╯
╭─ Next Steps ───────────────────────────────────────╮ │ - agents plan execute 01HXR5E6F7G8H9J0K1L2M3N4O5 │ │ - agents plan status 01HXR5E6F7G8H9J0K1L2M3N4O5 │ │ - agents plan tree 01HXR5E6F7G8H9J0K1L2M3N4O5 │ ╰────────────────────────────────────────────────────╯
✓ OK Plan created
The strategy produces a phased plan with child plans:
- Generate Alembic migration (add column with NULL default)
- Generate rollback migration
- Backfill from audit_log in batches of 10,000
- Update ORM model and application code
- Run tests with new schema
Each phase creates checkpoints. If the backfill fails at batch 850 of 1,000, the system can restore to the pre-backfill checkpoint rather than starting over.
Step 4: Monitor Execution and Rollback if Needed
# Monitor the migration execution $ agents plan status 01HXR5E6F7G8H9J0K1L2M3N4O5╭─ Plan Status ───────────────────────────────────────╮ │ Plan: 01HXR5E6F7G8H9J0K1L2M3N4O5 │ │ Phase: execute │ │ State: processing │ │ Action: local/add-column-with-backfill │ │ Project: local/api-service │ │ Automation: review │ │ Attempt: 1 │ ╰─────────────────────────────────────────────────────╯
╭─ Progress ─────────────────╮ │ ✓ Strategize │ │ ⏳ Execute (phase 3 of 5) │ │ • Apply (queued) │ ╰────────────────────────────╯
╭─ Child Plans ────────────────────────────────────────────────────────────╮ │ Phase Description Status │ │ ───── ─────────────────────────────────── ────────────────────── │ │ 1 Generate Alembic migration ✓ complete │ │ 2 Generate rollback migration ✓ complete │ │ 3 Backfill from audit_log ⏳ batch 620/1,000 │ │ 4 Update ORM model and application code ○ waiting │ │ 5 Run tests with new schema ○ waiting │ ╰──────────────────────────────────────────────────────────────────────────╯
╭─ Execution Detail ──────────────────────╮ │ Sandbox: transaction_rollback │ │ Tool Calls: 14 │ │ Files Modified: 2 │ │ Rows Backfilled: 6,200,000 / 10,000,000 │ │ Checkpoints: 4 created │ ╰─────────────────────────────────────────╯
╭─ Timing ──────────────╮ │ Started: 09:06:01 │ │ Elapsed: 00:04:32 │ │ ETA: 00:02:50 │ ╰───────────────────────╯
╭─ Cost ───────────────╮ │ Tokens Used: 18,940 │ │ Cost So Far: $0.063 │ │ Estimated: $0.110 │ ╰──────────────────────╯
✓ OK Status refreshed
# If something goes wrong during backfill, rollback to the pre-backfill checkpoint $ agents plan rollback --yes 01HXR5E6F7G8H9J0K1L2M3N4O5 cp_01HXR5CP3╭─ Rollback Summary ───────────────────────────────────╮ │ Plan: 01HXR5E6F7G8H9J0K1L2M3N4O5 │ │ Checkpoint: cp_01HXR5CP3 │ │ Label: before backfill (after migration applied) │ │ Files: 2 reverted │ ╰──────────────────────────────────────────────────────╯
╭─ Changes Reverted ──────────────────────────────────────────╮ │ File Action │ │ ─────────────────────────────────────────── ────────── │ │ alembic/versions/20260211_add_last_login.py restored │ │ alembic/versions/20260211_rollback.py restored │ ╰─────────────────────────────────────────────────────────────╯
╭─ Database Rollback ─────────────────────────────────────────╮ │ Transaction: rolled back │ │ Rows Affected: 6,200,000 updates reverted │ │ Schema State: restored to pre-backfill │ ╰─────────────────────────────────────────────────────────────╯
╭─ Impact ──────────────────────────────────────╮ │ Child Plans Invalidated: 1 (backfill phase) │ │ Sandbox: restored to cp_01HXR5CP3 │ │ Decisions After CP: 3 discarded │ │ Tool Calls After CP: 8 undone │ ╰───────────────────────────────────────────────╯
╭─ Post-Rollback State ──────────╮ │ Phase: execute │ │ State: queued (awaiting input) │ │ Checkpoints Remaining: 3 │ ╰────────────────────────────────╯
✓ OK Rollback complete
Example 6: Writing Technical Documentation from Codebase Analysis
Scenario: A project has grown to 50,000 lines of code with minimal documentation. The team wants CleverAgents to generate comprehensive API documentation, architecture guides, and developer onboarding docs by analyzing the actual codebase.
Complexity: Intermediate. Read-only action, large context usage, no sandbox modifications to source code.
Step 1: Create a Read-Only Documentation Action
Create actions/generate-docs.yaml:
name: local/generate-docs
description: "Generate comprehensive documentation from codebase analysis"
long_description: |
Analyze the project codebase to produce developer documentation:
API reference, architecture overview, module guides, and onboarding
material. Uses code intelligence (search, vector similarity, graph
queries) to understand structure and relationships.
definition_of_done: |
- API reference covers all public endpoints/functions
- Architecture document shows module dependencies and data flow
- Each major module has a developer guide
- An onboarding document provides setup and contribution instructions
- All generated docs are accurate to the current codebase
- Documentation is written in Markdown format
strategy_actor: anthropic/claude-3.5-sonnet
execution_actor: anthropic/claude-3.5-sonnet
read_only: false
reusable: true
state: available
args:
- name: doc_types
type: string
required: true
description: "Comma-separated list: api-reference, architecture, module-guides, onboarding"
- name: output_dir
type: string
required: false
default: "docs/"
description: "Output directory for generated docs"
invariants:
- "Do not modify any source code files — only create or update files in the output directory"
- "All code examples in documentation must be actual code from the project, not fabricated"
- "Architecture diagrams must reflect actual module dependencies, not aspirational ones"
$ agents action create --config ./actions/generate-docs.yaml╭─ Action Created ──────────────────────────────────────────────────────────────────╮ │ Name: local/generate-docs │ │ State: available │ │ Strategy Actor: anthropic/claude-3.5-sonnet │ │ Execution Actor: anthropic/claude-3.5-sonnet │ │ Reusable: yes │ │ Read Only: no │ │ Created: 2026-02-11 10:30 │ ╰───────────────────────────────────────────────────────────────────────────────────╯
╭─ Definition of Done ──────────────────────────────────────────────────────────────╮ │ - API reference covers all public endpoints/functions │ │ - Architecture document shows module dependencies and data flow │ │ - Each major module has a developer guide │ │ - An onboarding document provides setup and contribution instructions │ │ - All generated docs are accurate to the current codebase │ │ - Documentation is written in Markdown format │ ╰───────────────────────────────────────────────────────────────────────────────────╯
╭─ Arguments ───────────────────────────────────────────────────────────────────────────────────────╮ │ Name Type Required Description │ │ ────────── ────── ──────── ────────────────────────────────────────────────────────────── │ │ doc_types string yes Comma-separated list: api-reference, architecture, module-guides… │ │ output_dir string no Output directory for generated docs (default: docs/) │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Invariants ───────────────────────────────────────────────────────────────────────────────────────╮ │ 1. Do not modify any source code files — only create or update files in the output directory │ │ 2. All code examples in documentation must be actual code from the project, not fabricated │ │ 3. Architecture diagrams must reflect actual module dependencies, not aspirational ones │ ╰────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Automation ──────────────────────────╮ │ Profile: (inherits global) │ │ Source: default │ ╰───────────────────────────────────────╯
╭─ Usage ────────────────────────────────────────────────────────────────────────────╮ │ agents plan use local/generate-docs local/api-service │ │ --arg doc_types="api-reference,architecture,module-guides,onboarding" │ ╰────────────────────────────────────────────────────────────────────────────────────╯
✓ OK Action created
Step 2: Configure Context for Large Codebase Analysis
# Set a generous context policy for the strategize view (needs to see architecture) $ agents project context set --view strategize \ --include-resource 01HXR6A1B2C3D4E5F6G7H8J9K0 \ --exclude-path "**/node_modules/**" \ --exclude-path "**/.git/**" \ --exclude-path "**/dist/**" \ --hot-max-tokens 32000 \ --warm-max-decisions 200 \ --query-limit 50 \ --max-file-size 2097152 \ --summarize \ --summary-max-tokens 1500 \ local/api-service╭─ Context Policy ────────────────────────────────────────╮ │ Project: local/api-service │ │ View: strategize │ │ Include: 01HXR6A1B2C3D4E5F6G7H8J9K0 │ │ Exclude: /node_modules/, /.git/, /dist/ │ ╰─────────────────────────────────────────────────────────╯
╭─ Limits ──────────────────────────╮ │ Hot Tokens: 32000 (soft cap) │ │ Warm Decisions: 200 │ │ Query Limit: 50 │ │ Max File Size: 2 MB │ │ Max Total Size: 50 MB │ ╰───────────────────────────────────╯
╭─ Summarization ──────╮ │ Enabled: yes │ │ Max Tokens: 1500 │ ╰──────────────────────╯
╭─ Other Views ──────────╮ │ execute: (default) │ │ apply: (default) │ │ default: (unset) │ ╰────────────────────────╯
✓ OK Context policy updated
# Execution view can be narrower (writing docs, not analyzing all code) $ agents project context set --view execute \ --include-resource 01HXR6A1B2C3D4E5F6G7H8J9K0 \ --hot-max-tokens 16000 \ --query-limit 30 \ --summarize \ local/api-service╭─ Context Policy ────────────────────────────────────────╮ │ Project: local/api-service │ │ View: execute │ │ Include: 01HXR6A1B2C3D4E5F6G7H8J9K0 │ ╰─────────────────────────────────────────────────────────╯
╭─ Limits ──────────────────────────╮ │ Hot Tokens: 16000 (soft cap) │ │ Query Limit: 30 │ │ Max File Size: 1 MB │ │ Max Total Size: 50 MB │ ╰───────────────────────────────────╯
╭─ Summarization ─╮ │ Enabled: yes │ │ Max Tokens: 800 │ ╰─────────────────╯
╭─ Other Views ──────────────╮ │ strategize: configured │ │ apply: (default) │ │ default: (unset) │ ╰────────────────────────────╯
✓ OK Context policy updated
Step 3: Execute
$ agents plan use \ --automation-profile trusted \ --arg doc_types="api-reference,architecture,module-guides,onboarding" \ --arg output_dir="docs/generated/" \ local/generate-docs local/api-service╭─ Plan Created ──────────────────────────────────────╮ │ Plan: 01HXR6F7G8H9J0K1L2M3N4O5P6 │ │ Phase: strategize │ │ State: processing │ │ Action: local/generate-docs │ │ Project: local/api-service │ │ Automation: trusted │ │ Attempt: 1 │ ╰─────────────────────────────────────────────────────╯
╭─ Inputs ─────────────────────────────────────────────────────────────────────╮ │ - doc_types="api-reference,architecture,module-guides,onboarding" │ │ - output_dir="docs/generated/" │ │ - automation_profile=trusted │ ╰──────────────────────────────────────────────────────────────────────────────╯
╭─ Actors ────────────────────────────────────────╮ │ Strategy: anthropic/claude-3.5-sonnet │ │ Execution: anthropic/claude-3.5-sonnet │ │ Estimation: (none) │ ╰─────────────────────────────────────────────────╯
╭─ Automation ───────────────────────────╮ │ Profile: trusted │ │ Source: plan override │ │ Read-Only: no │ ╰────────────────────────────────────────╯
╭─ Context ──────────────────────────╮ │ Resources: 1 (git-checkout) │ │ Indexed Files: 487 │ │ View: strategize │ │ Hot Token Budget: 32,000 │ ╰────────────────────────────────────╯
╭─ Next Steps ─────────────────────────────────────────────────────────╮ │ Trusted profile — strategize and execute will proceed automatically. │ │ - agents plan status 01HXR6F7G8H9J0K1L2M3N4O5P6 │ │ - agents plan tree 01HXR6F7G8H9J0K1L2M3N4O5P6 │ ╰──────────────────────────────────────────────────────────────────────╯
✓ OK Plan created
The system analyzes the codebase using the code intelligence indexes (full-text, vector, and graph) to understand module relationships, public APIs, and data flows. It generates documentation files in the sandbox.
# Review generated docs $ agents plan artifacts 01HXR6F7G8H9J0K1L2M3N4O5P6╭─ Artifacts ───────────────────────────────────────────────────────────────────────────╮ │ Path Type Size Change Child Plan │ │ ──────────────────────────────────────── ───── ─────── ──────── ─────── │ │ docs/generated/api-reference.md write 18.2 KB +412 -0 root │ │ docs/generated/architecture.md write 12.6 KB +284 -0 root │ │ docs/generated/modules/auth.md write 6.4 KB +148 -0 root │ │ docs/generated/modules/billing.md write 5.8 KB +131 -0 root │ │ docs/generated/modules/notifications.md write 4.2 KB +97 -0 root │ │ docs/generated/modules/data-pipeline.md write 5.1 KB +118 -0 root │ │ docs/generated/onboarding.md write 8.9 KB +201 -0 root │ ╰───────────────────────────────────────────────────────────────────────────────────────╯
╭─ Summary ─────────────╮ │ Total: 7 │ │ Writes: 7 (new) │ │ Edits: 0 │ │ Deletes: 0 │ │ Total Size: 61.2 KB │ ╰───────────────────────╯
✓ OK 7 artifacts listed
$ agents plan diff 01HXR6F7G8H9J0K1L2M3N4O5P6╭─ Diff Summary ─────────────────────────────────────╮ │ Plan: 01HXR6F7G8H9J0K1L2M3N4O5P6 │ │ Project: local/api-service │ │ Files Changed: 7 │ │ Insertions: 1,391 │ │ Deletions: 0 │ │ Net Change: +1,391 lines │ ╰────────────────────────────────────────────────────╯
╭─ Files ──────────────────────────────────────────────────────────╮ │ Path Change Status │ │ ──────────────────────────────────────── ───────── ───────── │ │ docs/generated/api-reference.md +412 -0 new file │ │ docs/generated/architecture.md +284 -0 new file │ │ docs/generated/modules/auth.md +148 -0 new file │ │ docs/generated/modules/billing.md +131 -0 new file │ │ docs/generated/modules/notifications.md +97 -0 new file │ │ docs/generated/modules/data-pipeline.md +118 -0 new file │ │ docs/generated/onboarding.md +201 -0 new file │ ╰──────────────────────────────────────────────────────────────────╯
╭─ Patch Preview ───────────────────────────────────────────────────────╮ │ --- /dev/null │ │ +++ b/docs/generated/api-reference.md │ │ @@ -0,0 +1,412 @@ │ │ + # API Reference │ │ + ## Authentication │ │ + ### POST /auth/login │ │ + ... │ │ --- /dev/null │ │ +++ b/docs/generated/architecture.md │ │ @@ -0,0 +1,284 @@ │ │ + # Architecture Overview │ │ + ## Module Dependency Graph │ │ + ... │ ╰───────────────────────────────────────────────────────────────────────╯
╭─ Risk Assessment ─────────────────╮ │ Source Code: unmodified │ │ Breaking Changes: none detected │ │ New Files Only: yes │ ╰───────────────────────────────────╯
✓ OK Diff generated
# Apply the documentation to the project $ agents plan apply --yes 01HXR6F7G8H9J0K1L2M3N4O5P6╭─ Apply Summary ──────────────────────────────╮ │ Plan: 01HXR6F7G8H9J0K1L2M3N4O5P6 │ │ Artifacts: 7 files created │ │ Changes: 1,391 insertions, 0 deletions │ │ Project: local/api-service │ │ Applied At: 2026-02-11 10:38 │ ╰──────────────────────────────────────────────╯
╭─ Validation ────────────────────────────────╮ │ ✓ Invariant: no source files modified │ │ ✓ Invariant: code examples verified │ │ ✓ Invariant: architecture matches codebase │ │ Duration: 4.1s │ ╰─────────────────────────────────────────────╯
╭─ Sandbox Cleanup ─────────╮ │ Worktree: removed │ │ Branch: merged to main │ │ Checkpoint: archived │ ╰───────────────────────────╯
╭─ Plan Lifecycle ─────────────────────────╮ │ Phase: apply │ │ State: applied │ │ Total Duration: 00:08:12 │ │ Total Cost: $0.184 │ │ Decisions Made: 18 │ │ Child Plans: 0 │ ╰──────────────────────────────────────────╯
╭─ Next Steps ──────╮ │ - Review git diff │ │ - Commit changes │ ╰───────────────────╯
✓ OK Changes applied
Example 7: CI/CD Integration — Automated PR Review and Fix
Scenario: A CI pipeline triggers CleverAgents to automatically review incoming pull requests, run analysis, and if issues are found, create fix commits. This runs headlessly without human intervention.
Complexity: Advanced. Non-interactive (CI profile), headless execution, full automation.
Step 1: Configure for CI
# Set CI-appropriate configuration$ agents config set core.automation-profile ci
╭─ Config Updated ───────────────────────────╮ │ Key: core.automation-profile │ │ Value: ci │ │ Previous: (not set) │ │ Source: config │ │ Scope: global │ ╰────────────────────────────────────────────╯
╭─ Effective ────────────────────────╮ │ Sessions: new sessions │ │ Plans: future plans (unless set) │ │ Existing: unchanged │ ╰────────────────────────────────────╯
╭─ Saved To ──────────────────────────╮ │ File: ~/.cleveragents/config.toml │ │ Line: 3 │ ╰─────────────────────────────────────╯
✓ OK Config updated
$ agents config set core.format json
╭─ Config Updated ─────────────────╮ │ Key: core.format │ │ Value: json │ │ Previous: (not set) │ │ Source: config │ │ Scope: global │ ╰──────────────────────────────────╯
╭─ Saved To ──────────────────────────╮ │ File: ~/.cleveragents/config.toml │ │ Line: 4 │ ╰─────────────────────────────────────╯
✓ OK Config updated
$ agents config set core.log.level WARN
╭─ Config Updated ──────────────────╮ │ Key: core.log.level │ │ Value: WARN │ │ Previous: FATAL │ │ Source: config │ │ Scope: global │ ╰───────────────────────────────────╯
╭─ Saved To ──────────────────────────╮ │ File: ~/.cleveragents/config.toml │ │ Line: 5 │ ╰─────────────────────────────────────╯
✓ OK Config updated
Step 2: Create the PR Review Action
Create actions/review-pr.yaml:
name: local/review-pr
description: "Automatically review a PR and fix issues"
definition_of_done: |
- All lint issues are resolved
- Type checking passes
- Test coverage does not decrease
- Security scan passes
- All fixes are committed to the PR branch
strategy_actor: anthropic/claude-3.5-sonnet
execution_actor: anthropic/claude-3.5-sonnet
automation_profile: ci
reusable: true
state: available
args:
- name: pr_branch
type: string
required: true
description: "Branch name of the PR"
- name: base_branch
type: string
required: false
default: "main"
description: "Base branch to compare against"
invariants:
- "Only modify files that are already changed in the PR"
- "Do not change the intent of any code — only fix style, types, and test issues"
- "All fixes must include a comment explaining what was changed and why"
$ agents action create --config ./actions/review-pr.yaml╭─ Action Created ──────────────────────────────╮ │ Name: local/review-pr │ │ State: available │ │ Strategy Actor: anthropic/claude-3.5-sonnet │ │ Execution Actor: anthropic/claude-3.5-sonnet │ │ Reusable: yes │ │ Read Only: no │ │ Created: 2026-02-11 14:00 │ ╰───────────────────────────────────────────────╯
╭─ Definition of Done ─────────────────────────────────────╮ │ - All lint issues are resolved │ │ - Type checking passes │ │ - Test coverage does not decrease │ │ - Security scan passes │ │ - All fixes are committed to the PR branch │ ╰──────────────────────────────────────────────────────────╯
╭─ Arguments ──────────────────────────────────────────────────────╮ │ Name Type Required Description │ │ ────────── ────── ──────── ────────────────────────────────── │ │ pr_branch string yes Branch name of the PR │ │ base_branch string no Base branch to compare against │ ╰──────────────────────────────────────────────────────────────────╯
╭─ Invariants ─────────────────────────────────────────────────────────────────────╮ │ 1. Only modify files that are already changed in the PR │ │ 2. Do not change the intent of any code — only fix style, types, and test │ │ issues │ │ 3. All fixes must include a comment explaining what was changed and why │ ╰──────────────────────────────────────────────────────────────────────────────────╯
╭─ Automation ──────────────────╮ │ Profile: ci │ │ Source: action config │ ╰───────────────────────────────╯
╭─ Usage ──────────────────────────────────────────────────────────────────╮ │ agents plan use local/review-pr <PROJECT> │ │ --arg pr_branch="..." │ ╰──────────────────────────────────────────────────────────────────────────╯
✓ OK Action created
Step 3: CI Pipeline Script
#!/bin/bash # ci-review.sh — called by CI when a PR is opened or updatedset -euo pipefail
PR_BRANCH="${1:?Usage: ci-review.sh <branch>}" BASE_BRANCH="${2:-main}"
# Register the resource pointing to the CI workspace. RESOURCE_NAME="local/ci-${PR_BRANCH}" agents resource add git-checkout "$RESOURCE_NAME"
--path "$GITHUB_WORKSPACE"
--branch "$PR_BRANCH"echo "Registered resource: $RESOURCE_NAME"
# Create a project for this CI run, linking the resource by name. # If the project already exists from a previous run, ignore the error. agents --format json project create -d "CI workspace for PR $PR_BRANCH"
--resource "$RESOURCE_NAME"
local/ci-workspace 2>/dev/null || true# Register and attach validations (idempotent — duplicates are ignored) agents --format json validation add
--config validations/lint.yaml --required local/ci-lint 2>/dev/null || true agents --format json validation add
--config validations/typecheck.yaml --required local/ci-typecheck 2>/dev/null || true agents --format json validation add
--config validations/tests.yaml --required local/ci-tests 2>/dev/null || true agents --format json validation attach
--project local/ci-workspace local/ci-lint 2>/dev/null || true agents --format json validation attach
--project local/ci-workspace local/ci-typecheck 2>/dev/null || true agents --format json validation attach
--project local/ci-workspace local/ci-tests 2>/dev/null || true# Run the plan — ci profile means full automation including apply PLAN_OUTPUT=$(agents --format json plan use
--automation-profile ci
--arg pr_branch="$PR_BRANCH"
--arg base_branch="$BASE_BRANCH"
local/review-pr local/ci-workspace)PLAN_ID=$(echo "$PLAN_OUTPUT" | jq -r '.plan_id') echo "Plan started: $PLAN_ID"
# Wait for completion (ci profile runs everything automatically) while true; do STATUS=$(agents --format json plan status "$PLAN_ID" | jq -r '.state') case "$STATUS" in applied) echo "PR review complete. Changes applied." break ;; failed|cancelled) echo "PR review failed." agents --format json plan status "$PLAN_ID" exit 1 ;; *) sleep 10 ;; esac done
# Output the diff for the CI log agents --format json plan diff "$PLAN_ID"
Sample --format json output from resource add (captured into $RESOURCE_JSON):
{
"id": "01HXR7G1A2B3C4D5E6F7H8J9K0",
"type": "git-checkout",
"physical": true,
"path": "/home/runner/work/acme-api/acme-api",
"branch": "fix/handle-null-users",
"created": "2026-02-11T14:05:32Z",
"children_discovered": 247,
"capabilities": {
"readable": true,
"writable": true,
"sandboxable": true,
"checkpointable": true,
"sandbox_strategy": "git_worktree"
}
}
Sample --format json output from plan use (captured into $PLAN_OUTPUT):
{
"plan_id": "01HXR7H2B3C4D5E6F7G8J9K0L1",
"phase": "strategize",
"action": "local/review-pr",
"project": "local/ci-workspace",
"automation_profile": "ci",
"attempt": 1,
"args": {
"pr_branch": "fix/handle-null-users",
"base_branch": "main"
},
"resources": ["01HXR7G1A2B3C4D5E6F7H8J9K0"]
}
Example 8: Cloud Infrastructure Management
Scenario: A DevOps team manages Terraform infrastructure. They want CleverAgents to analyze their current infrastructure, propose optimizations (right-sizing, unused resources), and generate the Terraform changes.
Complexity: Advanced. Custom resource types, custom skills, infrastructure domain.
Step 1: Register Custom Resource Types and Skills
Create resource-types/terraform-state.yaml:
name: local/terraform-state
description: "Terraform state file resource type"
physical: true
user_addable: true
sandbox_strategy: filesystem_copy
cli_args:
- name: state-path
type: string
required: true
description: "Path to terraform.tfstate or remote state configuration"
- name: workspace
type: string
required: false
default: "default"
description: "Terraform workspace name"
handler: local/terraform-state-handler
child_types:
- type: fs-file
relationship: contains
Create skills/terraform-ops.yaml:
name: local/terraform-ops
description: "Terraform infrastructure operations"
tools:
- name: terraform_plan
description: "Run terraform plan and return the execution plan"
input_schema:
type: object
properties:
target:
type: string
description: "Specific resource to target (optional)"
required: []
writes: false
checkpointable: false
-
name: terraform_show
description: "Show current state of a Terraform resource"
input_schema:
type: object
properties:
resource_address:
type: string
description: "Terraform resource address (e.g., aws_instance.web)"
required: [resource_address]
writes: false
checkpointable: false
-
name: cloud_metrics
description: "Fetch CloudWatch/cloud monitoring metrics for a resource"
input_schema:
type: object
properties:
resource_id:
type: string
metric:
type: string
description: "e.g., CPUUtilization, NetworkIn, RequestCount"
period_days:
type: integer
default: 30
required: [resource_id, metric]
writes: false
checkpointable: false
include_skills:
local/file-ops
$ agents resource type add --config ./resource-types/terraform-state.yaml╭─ Resource Type ──────────────────────────────╮ │ Name: local/terraform-state │ │ Physical/Virtual: physical │ │ User Addable: yes │ │ Registered: 2026-02-11 14:30 │ ╰──────────────────────────────────────────────╯
╭─ CLI Arguments ──────────────────────────────────────────────────────────────╮ │ Argument Required Description │ │ ────────────── ──────── ────────────────────────────────────────────── │ │ --state-path yes Path to terraform.tfstate or remote state config │ │ --workspace no Terraform workspace name (default: "default") │ ╰──────────────────────────────────────────────────────────────────────────────╯
╭─ Child Types ──────────────╮ │ Auto-discover: fs-file │ ╰────────────────────────────╯
╭─ Sandbox ───────────────────────────────────────────╮ │ Strategy: filesystem_copy │ │ Handler: local/terraform-state-handler (custom) │ ╰─────────────────────────────────────────────────────╯
✓ OK Resource type registered ℹ New subcommand available: agents resource add local/terraform-state
$ agents skill add --config ./skills/terraform-ops.yaml╭─ Skill ──────────────────────────────────────────╮ │ Name: local/terraform-ops │ │ Description: Terraform infrastructure operations │ │ Config: ./skills/terraform-ops.yaml │ │ Created: 2026-02-11 14:31 │ ╰──────────────────────────────────────────────────╯
╭─ Includes ────────────────────╮ │ local/file-ops (registered) │ ╰───────────────────────────────╯
╭─ Tool Sources ───────────────────────────────────────────────────────╮ │ Source Count Details │ │ ──────── ───── ────────────────────────────────────────────── │ │ custom 3 terraform_plan, terraform_show, cloud_metrics │ │ included 5 local/file-ops (5) │ │ ──────── ───── ────────────────────────────────────────────── │ │ Total: 8 │ ╰──────────────────────────────────────────────────────────────────────╯
✓ OK Skill registered with 8 tools
Step 2: Set Up the Project
# Register the infrastructure git checkout $ agents resource add git-checkout local/infra-repo \ --path /repos/infrastructure --branch main╭─ Resource ──────────────────────────────────────╮ │ Name: local/infra-repo │ │ ID: 01HXR8A1B2C3D4E5F6G7H8J9K0 │ │ Type: git-checkout │ │ Physical/Virtual: physical │ │ Path: /repos/infrastructure │ │ Branch: main │ │ Created: 2026-02-11 14:32 │ ╰─────────────────────────────────────────────────╯
╭─ Auto-discovered Children ─────────────────────────────────────────╮ │ ID Type Status │ │ ─────────────── ────────────── ───────────────── │ │ 01HXR8A1B2C3… git created │ │ 01HXR8A1B2C4… git-remote created │ │ 01HXR8A1B2C5… git-branch created │ │ 01HXR8A1B2C6… fs-directory created │ │ + 18 git-commit resources │ │ + 84 git-tree-entry resources │ │ + 6 fs-directory + 42 fs-file │ ╰────────────────────────────────────────────────────────────────────╯
╭─ Capabilities ─────────────────╮ │ Readable: yes │ │ Writable: yes │ │ Sandboxable: yes │ │ Checkpointable: yes │ │ Sandbox Strategy: git_worktree │ ╰────────────────────────────────╯
✓ OK Resource registered (155 child resources discovered)
# Register the Terraform state for the production workspace $ agents resource add local/terraform-state local/prod-tfstate \ --state-path /repos/infrastructure/environments/prod/terraform.tfstate \ --workspace production╭─ Resource ──────────────────────────────────────────────────────────────────╮ │ Name: local/prod-tfstate │ │ ID: 01HXR8B2C3D4E5F6G7H8J9K0L1 │ │ Type: local/terraform-state │ │ Physical/Virtual: physical │ │ State Path: /repos/infrastructure/environments/prod/terraform.tfstate │ │ Workspace: production │ │ Created: 2026-02-11 14:32 │ ╰─────────────────────────────────────────────────────────────────────────────╯
╭─ Auto-discovered Children ─────────────╮ │ + 1 fs-file (terraform.tfstate) │ ╰────────────────────────────────────────╯
╭─ Capabilities ──────────────────────╮ │ Readable: yes │ │ Writable: yes │ │ Sandboxable: yes │ │ Checkpointable: yes │ │ Sandbox Strategy: filesystem_copy │ ╰─────────────────────────────────────╯
✓ OK Resource registered (1 child resource discovered)
# Create the project, linking both resources by name $ agents project create -d "Production infrastructure" \ --resource local/infra-repo \ --resource local/prod-tfstate \ --invariant "Never delete or modify resources tagged with 'critical: true'" \ --invariant "All changes must be backward-compatible with running services" \ --invariant "Cost changes must be documented in the PR description" \ local/infra-prod╭─ Project Created ─────────────────────────────╮ │ Name: local/infra-prod │ │ Description: Production infrastructure │ │ Remote: no │ ╰───────────────────────────────────────────────╯
╭─ Linked Resources ──────────────────────────────────────────────────╮ │ Resource Type Read-Only │ │ ──────────────────────────── ────────────────────── ───────── │ │ local/infra-repo git-checkout no │ │ local/prod-tfstate local/terraform-state no │ ╰─────────────────────────────────────────────────────────────────────╯
╭─ Invariants ──────────────────────────────────────────────────────────────╮ │ 1. Never delete or modify resources tagged with 'critical: true' │ │ 2. All changes must be backward-compatible with running services │ │ 3. Cost changes must be documented in the PR description │ ╰───────────────────────────────────────────────────────────────────────────╯
✓ OK Project created
$ agents validation add --config validations/tf-validate.yaml --required local/tf-validate╭─ Validation Registered ────────────────────────────────────────────╮ │ Name: local/tf-validate │ │ Command: cd environments/prod && terraform validate │ │ Mode: required │ │ Timeout: 300s │ ╰────────────────────────────────────────────────────────────────────╯
✓ OK Validation registered
$ agents validation attach --project local/infra-prod local/tf-validate ✓ OK Validation attached to project local/infra-prod
$ agents validation add --config validations/tf-plan.yaml --informational local/tf-plan╭─ Validation Registered ─────────────────────────────────────────────────────────╮ │ Name: local/tf-plan │ │ Command: cd environments/prod && terraform plan -detailed-exitcode │ │ Mode: informational │ │ Timeout: 300s │ ╰─────────────────────────────────────────────────────────────────────────────────╯
✓ OK Validation registered
$ agents validation attach --project local/infra-prod local/tf-plan ✓ OK Validation attached to project local/infra-prod
Step 3: Create and Run the Optimization Action
Create actions/infra-optimize.yaml:
name: local/infra-optimize
description: "Analyze and optimize cloud infrastructure costs"
definition_of_done: |
- Unused resources are identified and removal plans generated
- Over-provisioned resources are identified with right-sizing recommendations
- Terraform changes implement the approved optimizations
- terraform validate passes on all changes
- terraform plan shows no unexpected resource deletions
- Estimated monthly cost savings are documented
strategy_actor: local/refactoring-strategist
execution_actor: anthropic/claude-3.5-sonnet
automation_profile: supervised
reusable: true
state: available
args:
- name: optimization_targets
type: string
required: false
default: "all"
description: "Comma-separated: compute, storage, networking, database, all"
- name: min_savings_threshold
type: float
required: false
default: 10.0
description: "Minimum monthly savings (USD) to include an optimization"
invariants:
- "Changes to production databases require manual approval regardless of profile"
- "Instance type downgrades must verify current peak CPU < 70% of new capacity"
$ agents action create --config ./actions/infra-optimize.yaml╭─ Action Created ──────────────────────────────────────╮ │ Name: local/infra-optimize │ │ ID: 01HXR8C3D4E5F6G7H8J9K0L1M2 │ │ State: available │ │ Strategy Actor: local/refactoring-strategist │ │ Execution Actor: anthropic/claude-3.5-sonnet │ │ Reusable: yes │ │ Read Only: no │ │ Config: ./actions/infra-optimize.yaml │ │ Created: 2026-02-11 14:33 │ ╰───────────────────────────────────────────────────────╯
╭─ Definition of Done ──────────────────────────────────────────────────╮ │ - Unused resources are identified and removal plans generated │ │ - Over-provisioned resources identified with right-sizing recs │ │ - Terraform changes implement the approved optimizations │ │ - terraform validate passes on all changes │ │ - terraform plan shows no unexpected resource deletions │ │ - Estimated monthly cost savings are documented │ ╰───────────────────────────────────────────────────────────────────────╯
╭─ Arguments ───────────────────────────────────────────────────────────────────────╮ │ Name Type Required Description │ │ ────────────────────── ────── ──────── ────────────────────────────────── │ │ optimization_targets string no Comma-separated: compute, storage, … │ │ min_savings_threshold float no Min monthly savings (USD) to include │ ╰───────────────────────────────────────────────────────────────────────────────────╯
╭─ Invariants ──────────────────────────────────────────────────────────────────────────╮ │ 1. Changes to production databases require manual approval regardless of profile │ │ 2. Instance type downgrades must verify current peak CPU < 70% of new capacity │ ╰───────────────────────────────────────────────────────────────────────────────────────╯
╭─ Automation ──────────────────────────╮ │ Profile: supervised │ │ Source: action config │ ╰───────────────────────────────────────╯
╭─ Usage ─────────────────────────────────────────────────────────────╮ │ agents plan use local/infra-optimize local/infra-prod │ │ --arg optimization_targets=TARGET --arg min_savings_threshold=N │ ╰─────────────────────────────────────────────────────────────────────╯
✓ OK Action created
$ agents plan use \ --arg optimization_targets="compute,storage" \ --arg min_savings_threshold=25.0 \ local/infra-optimize local/infra-prod╭─ Plan Created ─────────────────────────────────────────╮ │ Plan ID: 01HXR8D4E5F6G7H8J9K0L1M2N3 │ │ Phase: strategize │ │ Action: local/infra-optimize │ │ Project: local/infra-prod │ │ Automation: supervised │ │ Attempt: 1 │ ╰────────────────────────────────────────────────────────╯
╭─ Inputs ──────────────────────────────────────╮ │ - optimization_targets=compute,storage │ │ - min_savings_threshold=25.0 │ ╰───────────────────────────────────────────────╯
╭─ Actors ─────────────────────────────────────╮ │ Strategy: local/refactoring-strategist │ │ Execution: anthropic/claude-3.5-sonnet │ │ Estimation: (none) │ ╰──────────────────────────────────────────────╯
╭─ Plan Invariants ──────────────────────────────────────────────────────────────────╮ │ Scope Source Invariant │ │ ──────── ──────── ───────────────────────────────────────────────────────── │ │ plan action Prod DB changes require manual approval regardless │ │ plan action Instance downgrades must verify peak CPU < 70% of new cap │ │ project config Never delete or modify resources tagged 'critical: true' │ │ project config All changes must be backward-compatible with running svcs │ │ project config Cost changes must be documented in PR description │ ╰────────────────────────────────────────────────────────────────────────────────────╯
╭─ Context ──────────────────────────────────────────╮ │ Resources: 2 (git-checkout, terraform-state) │ │ Indexed Files: 156 │ │ View: strategize │ │ Hot Token Budget: 12,000 │ ╰────────────────────────────────────────────────────╯
╭─ Next Steps ──────────────────────────────────────────╮ │ - agents plan execute 01HXR8D4E5F6G7H8J9K0L1M2N3 │ │ - agents plan status 01HXR8D4E5F6G7H8J9K0L1M2N3 │ │ - agents plan tree 01HXR8D4E5F6G7H8J9K0L1M2N3 │ ╰───────────────────────────────────────────────────────╯
✓ OK Plan created
The strategy actor queries cloud metrics via the cloud_metrics tool, analyzes the Terraform state using terraform_show, and proposes optimizations with cost estimates. Because the automation profile is supervised, the engineer reviews the full strategy — including any proposed instance downgrades or resource removals — before execution modifies any Terraform files. The execution actor then generates the .tf changes, and both validations (terraform validate and terraform plan -detailed-exitcode) run automatically before the plan can be applied.
Example 9: Session-Driven Interactive Exploration
Scenario: A developer uses a CleverAgents session to interactively explore a new codebase, ask questions, and gradually build up an action to automate a recurring task.
Complexity: Beginner-Intermediate. Session-based, conversational, action creation through natural language.
Step 1: Start a Session
$ agents session create --actor anthropic/claude-3.5-sonnet╭─ Session ────────────────────────────────╮ │ ID: 01HXR9A1B2C3D4E5F6G7H8I9J0 │ │ Actor: anthropic/claude-3.5-sonnet │ │ Created: 2026-02-11 09:15 │ │ Namespace: local │ ╰──────────────────────────────────────────╯
╭─ Settings ─────────────╮ │ Automation: review │ │ Streaming: off │ │ Context: default │ │ Memory: enabled │ │ Max History: 50 turns │ ╰────────────────────────╯
╭─ Actor Details ────────────────────────╮ │ Provider: anthropic │ │ Model: claude-3.5-sonnet │ │ Temperature: 0.7 │ │ Context Window: 200K tokens │ ╰────────────────────────────────────────╯
✓ OK Session created
Step 2: Explore the Codebase Interactively
$ agents session tell --session 01HXR9A1B2C3D4E5F6G7H8I9J0 \ "I just joined the team and need to understand the api-service project. What are the main modules and how do they interact?"╭─ Plan Request ──────────────────────────────────────────────────────────────────────────╮ │ Actor: anthropic/claude-3.5-sonnet │ │ Session: 01HXR9A1B2C3D4E5F6G7H8I9J0 │ │ Automation: review │ │ Prompt: I just joined the team and need to understand the api-service project. Wha... │ ╰─────────────────────────────────────────────────────────────────────────────────────────╯
▍ The api-service project is organized into 5 core modules:
src/api/routes/— HTTP route definitions (REST + GraphQL). Each domain entity has its own router file (e.g.users.py,orders.py,products.py).
src/api/services/— Business logic layer. Services are injected into routes via dependency injection. Key files:user_service.py,order_service.py,notification_service.py.
src/api/models/— SQLAlchemy ORM models and Pydantic schemas. The models map 1:1 to database tables; schemas handle request/response validation.
src/api/auth/— Authentication and authorization. Implements JWT-based auth with role-based access control. Middleware inauth_middleware.pyintercepts every request before it reaches a route handler.
src/api/workers/— Background task processors using Celery. Handles async jobs like email dispatch, report generation, and webhook delivery.Data flow: Routes → Services → Models → Database. Auth middleware gates every route. Workers are dispatched by services for async operations.
╭─ Commands Executed ──────────────────────────────────────────────────────────────╮ │ - agents resource inspect local/api-service-repo --tree │ │ - agents resource inspect local/api-service-repo --file src/api/__init__.py │ ╰──────────────────────────────────────────────────────────────────────────────────╯
╭─ Usage ─────────────────────╮ │ Input Tokens: 2,480 │ │ Output Tokens: 846 │ │ Cost: $0.0128 │ │ Duration: 4.1s │ │ Tool Calls: 2 │ ╰─────────────────────────────╯
✓ OK Orchestrator completed 2 commands
$ agents session tell --session 01HXR9A1B2C3D4E5F6G7H8I9J0 \ "Show me how authentication works. What's the flow from request to response?"╭─ Plan Request ───────────────────────────────────────────────────────────────────╮ │ Actor: anthropic/claude-3.5-sonnet │ │ Session: 01HXR9A1B2C3D4E5F6G7H8I9J0 │ │ Automation: review │ │ Prompt: Show me how authentication works. What's the flow from request to res... │ ╰──────────────────────────────────────────────────────────────────────────────────╯
▍ Here is the authentication flow in api-service:
1. Request arrives →
auth_middleware.py:34extracts theAuthorizationheader and parses the Bearer token.2. Token validation →
token_service.py:71decodes the JWT using the public key fromconfig.AUTH_PUBLIC_KEY. It verifies: - Signature (RS256) - Expiry (expclaim, default 3600s) - Issuer (issmust matchapi-service)3. User resolution →
user_service.py:45loads the user record from the database using thesubclaim. Returns 401 if the user is inactive or deleted.4. Permission check →
rbac.py:22evaluates the user's roles against the endpoint's required permissions (defined as decorators on each route, e.g.@require_role("admin")). Returns 403 on insufficient permissions.5. Context injection → The authenticated user object is attached to
request.state.userso downstream route handlers can access it without repeating lookups.Key files:
src/api/auth/auth_middleware.py,src/api/auth/token_service.py,src/api/auth/rbac.py,src/api/services/user_service.py.╭─ Commands Executed ──────────────────────────────────────────────────────────────────────╮ │ - agents resource inspect local/api-service-repo --file src/api/auth/auth_middleware.py │ │ - agents resource inspect local/api-service-repo --file src/api/auth/token_service.py │ │ - agents resource inspect local/api-service-repo --file src/api/auth/rbac.py │ ╰──────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Usage ─────────────────────╮ │ Input Tokens: 3,960 │ │ Output Tokens: 1,024 │ │ Cost: $0.0187 │ │ Duration: 5.3s │ │ Tool Calls: 3 │ ╰─────────────────────────────╯
✓ OK Orchestrator completed 3 commands
$ agents session tell --session 01HXR9A1B2C3D4E5F6G7H8I9J0 \ "I notice the team manually runs a dependency lock refresh every week. Can you create an action that automates this?"╭─ Plan Request ──────────────────────────────────────────────────────────────────────────╮ │ Actor: anthropic/claude-3.5-sonnet │ │ Session: 01HXR9A1B2C3D4E5F6G7H8I9J0 │ │ Automation: review │ │ Prompt: I notice the team manually runs a dependency lock refresh every week. Can... │ ╰─────────────────────────────────────────────────────────────────────────────────────────╯
▍ I've created a local/refresh-locks action that automates the weekly dependency lock refresh. Here's what it does:
- Checks out the target project's repository
- Runs
pip compile/poetry lock(auto-detected from project config)- Runs the test suite to verify nothing broke
- Commits the updated lock files on a new branch
- Opens a pull request for review
The action is now available. You can review the generated config with
agents action show local/refresh-locks.╭─ Commands Executed ──────────────────────────────────────────────────────╮ │ - agents action create --config ./actions/refresh-locks.yaml │ ╰──────────────────────────────────────────────────────────────────────────╯
╭─ Result ────────────────────────────╮ │ Action: local/refresh-locks │ │ Steps: 5 │ │ Trigger: manual │ ╰───────────────────────────────────────╯
╭─ Usage ─────────────────────╮ │ Input Tokens: 5,210 │ │ Output Tokens: 1,380 │ │ Cost: $0.0249 │ │ Duration: 6.7s │ │ Tool Calls: 1 │ ╰─────────────────────────────╯
✓ OK Orchestrator completed 1 command
Step 3: Review Session History
$ agents session show 01HXR9A1B2C3D4E5F6G7H8I9J0╭─ Session Summary ─────────────────────────────╮ │ ID: 01HXR9A1B2C3D4E5F6G7H8I9J0 │ │ Actor: anthropic/claude-3.5-sonnet │ │ Messages: 6 │ │ Created: 2026-02-11 09:15 │ │ Updated: 2026-02-11 09:28 │ │ Automation: review │ ╰───────────────────────────────────────────────╯
╭─ Recent Messages ──────────────────────────────────────────────────────────────────────╮ │ user I just joined the team and need to understand the api-service project... │ │ assistant The api-service project is organized into 5 core modules... │ │ user Show me how authentication works... │ │ assistant Here is the authentication flow in api-service... │ │ user I notice the team manually runs a dependency lock refresh... │ │ assistant I've created a local/refresh-locks action that automates... │ ╰────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Linked Plans ──────────────────────────────────╮ │ Plan ID Phase State │ │ ─────────────────────────── ─────── ──────── │ │ 01HXR9K4M7N2P3Q5R6S8T9U0V1 execute complete │ ╰─────────────────────────────────────────────────╯
╭─ Token Usage ──────────────╮ │ Input Tokens: 11,650 │ │ Output Tokens: 3,250 │ │ Estimated Cost: $0.0564 │ ╰────────────────────────────╯
✓ OK Session details loaded
$ agents session export --output /tmp/onboarding-session.json 01HXR9A1B2C3D4E5F6G7H8I9J0╭─ Session Export ────────────────────────────╮ │ Session: 01HXR9A1B2C3D4E5F6G7H8I9J0 │ │ Output: /tmp/onboarding-session.json │ │ Messages: 6 │ │ Size: 38 KB │ │ Format: JSON │ ╰─────────────────────────────────────────────╯
╭─ Contents ─────────────────╮ │ Messages: 6 │ │ Plan References: 1 │ │ Metadata Keys: 3 │ │ Actor Config: included │ │ Schema Version: v3 │ ╰────────────────────────────╯
╭─ Integrity ──────────────────╮ │ Checksum: sha256:3e8f...b7d2 │ │ Encrypted: no │ ╰──────────────────────────────╯
✓ OK Export completed
Example 10: Full-Auto Batch Operations — Formatting and Linting
Scenario: A team has 15 Python packages in a monorepo that need to be reformatted with ruff format and have their imports sorted. This is low-risk and can be fully automated.
Complexity: Simple. Full-auto profile, batch execution, minimal risk.
Step 1: Create a Formatting Action
Create actions/format-codebase.yaml:
name: local/format-codebase
description: "Apply code formatting and import sorting"
definition_of_done: |
- ruff format passes with zero changes remaining
- ruff check --select I passes (imports sorted)
- All existing tests pass
- No semantic changes to any code
strategy_actor: anthropic/claude-3.5-sonnet
execution_actor: anthropic/claude-3.5-sonnet
automation_profile: full-auto
reusable: true
state: available
args:
- name: formatter_config
type: string
required: false
default: "pyproject.toml"
description: "Path to the formatter configuration file"
invariants:
- "Only whitespace and import ordering changes — no semantic modifications"
- "Every package must pass its own test suite after formatting"
$ agents action create --config ./actions/format-codebase.yaml╭─ Action Created ─────────────────────────────────╮ │ Name: local/format-codebase │ │ State: available │ │ Strategy Actor: anthropic/claude-3.5-sonnet │ │ Execution Actor: anthropic/claude-3.5-sonnet │ │ Reusable: yes │ │ Read Only: no │ │ Created: 2026-02-11 14:00 │ ╰──────────────────────────────────────────────────╯
╭─ Definition of Done ──────────────────────────────────────╮ │ - ruff format passes with zero changes remaining │ │ - ruff check --select I passes (imports sorted) │ │ - All existing tests pass │ │ - No semantic changes to any code │ ╰───────────────────────────────────────────────────────────╯
╭─ Arguments ──────────────────────────────────────────────────────────────────╮ │ Name Type Required Description │ │ ──────────────── ────── ──────── ────────────────────────────────────── │ │ formatter_config string no Path to the formatter configuration file │ ╰──────────────────────────────────────────────────────────────────────────────╯
╭─ Invariants ──────────────────────────────────────────────────────────────────────╮ │ 1. Only whitespace and import ordering changes — no semantic modifications │ │ 2. Every package must pass its own test suite after formatting │ ╰───────────────────────────────────────────────────────────────────────────────────╯
╭─ Automation ──────────────────────────╮ │ Profile: full-auto │ │ Source: action │ ╰───────────────────────────────────────╯
╭─ Usage ─────────────────────────────────────────────────────────╮ │ agents plan use local/format-codebase local/<project> │ │ --arg formatter_config="pyproject.toml" (optional) │ ╰─────────────────────────────────────────────────────────────────╯
✓ OK Action created
Step 2: Register All Packages and Run in Batch
# Register all 15 packages as resources and projects. # resource add accepts a name — use it to link to project create.
for pkg in auth billing common gateway inventory notifications
orders payments reporting search shipping users webhooks
workers analytics; do agents resource add git-checkout "local/pkg-${pkg}-repo"
--path "/repos/monorepo/packages/${pkg}" --branch main agents project create -d "Package: ${pkg}"
--resource "local/pkg-${pkg}-repo"
"local/pkg-${pkg}" agents validation add --config "validations/pkg-${pkg}.yaml" --required "local/pkg-${pkg}-test" agents validation attach --project "local/pkg-${pkg}" "local/pkg-${pkg}-test" done
# Loop output — first 2 iterations shown, remaining 13 condensed# --- auth ---
$ agents resource add git-checkout local/pkg-auth-repo
--path /repos/monorepo/packages/auth --branch main╭─ Resource ──────────────────────────────────────────╮ │ Name: local/pkg-auth-repo │ │ ID: 01HXR7B1A2B3C4D5E6F7G8H9J0 │ │ Type: git-checkout │ │ Physical/Virtual: physical │ │ Path: /repos/monorepo/packages/auth │ │ Branch: main │ │ Created: 2026-02-11 14:01 │ ╰─────────────────────────────────────────────────────╯
╭─ Auto-discovered Children ─────────────────────────────────────────╮ │ ID Type Status │ │ ─────────────── ────────────── ───────────────── │ │ 01HXR7B1A2B3… git created │ │ 01HXR7B1A2B4… git-remote created │ │ 01HXR7B1A2B5… git-branch created │ │ + 22 git-commit resources │ │ + 134 git-tree-entry resources │ │ + 2 fs-directory + 11 fs-file │ ╰────────────────────────────────────────────────────────────────────╯
╭─ Capabilities ─────────────────╮ │ Readable: yes │ │ Writable: yes │ │ Sandboxable: yes │ │ Checkpointable: yes │ │ Sandbox Strategy: git_worktree │ ╰────────────────────────────────╯
✓ OK Resource registered (172 child resources discovered)
$ agents project create -d "Package: auth"
--resource local/pkg-auth-repo local/pkg-auth╭─ Project Created ──────────────────────────────╮ │ Name: local/pkg-auth │ │ Description: Package: auth │ │ Type: local │ │ Created: 2026-02-11 14:01 │ ╰────────────────────────────────────────────────╯
╭─ Linked Resources ───────────────────────────────────────╮ │ Resource Type Read-Only │ │ ───────────────── ────────────── ───────── │ │ local/pkg-auth-repo git-checkout no │ ╰──────────────────────────────────────────────────────────╯
╭─ Defaults ──────────────────────────────╮ │ Sandbox: git_worktree │ │ Validations: 0 │ │ Context Filters: none │ │ Automation Profile: (inherits global) │ ╰─────────────────────────────────────────╯
✓ OK Project created
$ agents validation add --config validations/pkg-auth.yaml --required local/pkg-auth-test
╭─ Validation Registered ─────────────────────────────────────────────────╮ │ Name: local/pkg-auth-test │ │ Command: cd /repos/monorepo/packages/auth && pytest tests/ │ │ Mode: required │ │ Timeout: 300s │ ╰─────────────────────────────────────────────────────────────────────────╯
✓ OK Validation registered
$ agents validation attach --project local/pkg-auth local/pkg-auth-test ✓ OK Validation attached to project local/pkg-auth
# --- billing ---
$ agents resource add git-checkout local/pkg-billing-repo
--path /repos/monorepo/packages/billing --branch main ✓ OK Resource registered — Name: local/pkg-billing-repo (189 children)$ agents project create -d "Package: billing"
--resource local/pkg-billing-repo local/pkg-billing ✓ OK Project created — Name: local/pkg-billing$ agents validation add --config validations/pkg-billing.yaml --required local/pkg-billing-test ✓ OK Validation registered — Name: local/pkg-billing-test
$ agents validation attach --project local/pkg-billing local/pkg-billing-test ✓ OK Validation attached to project local/pkg-billing
# ... (common, gateway, inventory, notifications, orders, payments, # reporting, search, shipping, users, webhooks, workers, analytics) # Each iteration: resource add → project create → validation add → validation attach # All 15 packages registered.
# Run formatting across all packages. # full-auto: Strategize, Execute, AND Apply all run without human intervention. # Capture plan IDs so we can check results afterward.
PLAN_IDS=() for pkg in auth billing common gateway inventory notifications
orders payments reporting search shipping users webhooks
workers analytics; do PID=$(agents --format json plan use
--automation-profile full-auto
local/format-codebase "local/pkg-${pkg}"
| jq -r '.data.plan_id') PLAN_IDS+=("$PID") done
# Plan creation output — first 2 shown, remaining condensed# --- auth ---
$ agents plan use
--automation-profile full-auto
local/format-codebase local/pkg-auth╭─ Plan Created ───────────────────────────────────╮ │ Plan ID: 01HXR7D1A2B3C4D5E6F7G8H9J0 │ │ Phase: strategize (running) │ │ Action: local/format-codebase │ │ Project: local/pkg-auth │ │ Automation: full-auto │ │ Attempt: 1 │ ╰──────────────────────────────────────────────────╯
╭─ Invariants ─────────────────────────────────────────────────────────────╮ │ Scope Source Invariant │ │ ────── ────── ───────────────────────────────────────────────────── │ │ plan action Only whitespace and import ordering changes — no se… │ │ plan action Every package must pass its own test suite after fo… │ ╰──────────────────────────────────────────────────────────────────────────╯
✓ OK Plan created — full-auto: strategize → execute → apply will proceed automatically
# --- billing ---
$ agents plan use
--automation-profile full-auto
local/format-codebase local/pkg-billing╭─ Plan Created ───────────────────────────────────╮ │ Plan ID: 01HXR7D2B3C4D5E6F7G8H9J0K1 │ │ Phase: strategize (running) │ │ Action: local/format-codebase │ │ Project: local/pkg-billing │ │ Automation: full-auto │ │ Attempt: 1 │ ╰──────────────────────────────────────────────────╯
✓ OK Plan created — full-auto: strategize → execute → apply will proceed automatically
# ... (common, gateway, inventory, notifications, orders, payments, # reporting, search, shipping, users, webhooks, workers, analytics) # 15 plans launched — all running full-auto in background.
# Wait for all plans to finish, then check results$ agents plan list --state applied
╭─ Plans ──────────────────────────────────────────────────────────────────────────────────────────────╮ │ ID Phase State Action Project Elapsed │ │ ──────── ─────── ──────── ───────────────────── ───────────────────── ───────── │ │ 01HXR7D1 apply applied local/format-codebase local/pkg-auth 00:02:14 │ │ 01HXR7D2 apply applied local/format-codebase local/pkg-billing 00:02:31 │ │ 01HXR7D3 apply applied local/format-codebase local/pkg-common 00:01:47 │ │ 01HXR7D4 apply applied local/format-codebase local/pkg-gateway 00:02:08 │ │ 01HXR7D5 apply applied local/format-codebase local/pkg-inventory 00:01:55 │ │ 01HXR7D6 apply applied local/format-codebase local/pkg-notifications 00:02:42 │ │ 01HXR7D7 apply applied local/format-codebase local/pkg-orders 00:02:19 │ │ 01HXR7D8 apply applied local/format-codebase local/pkg-payments 00:01:58 │ │ 01HXR7D9 apply applied local/format-codebase local/pkg-reporting 00:02:36 │ │ 01HXR7DA apply applied local/format-codebase local/pkg-search 00:01:41 │ │ 01HXR7DB apply applied local/format-codebase local/pkg-shipping 00:02:22 │ │ 01HXR7DC apply applied local/format-codebase local/pkg-users 00:01:49 │ │ 01HXR7DD apply applied local/format-codebase local/pkg-webhooks 00:02:05 │ │ 01HXR7DE apply applied local/format-codebase local/pkg-analytics 00:02:11 │ ╰──────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Filters ───────────────────╮ │ Phase: (any) │ │ State: applied │ │ Project: (any) │ │ Action: (any) │ ╰─────────────────────────────╯
╭─ Summary ─────────────╮ │ Total: 14 │ │ Processing: 0 │ │ Completed: 14 │ │ Errored: 0 │ ╰───────────────────────╯
✓ OK 14 plans listed
$ agents plan list --state failed
╭─ Plans ──────────────────────────────────────────────────────────────────────────────────────────────╮ │ ID Phase State Action Project Elapsed │ │ ──────── ─────── ─────── ───────────────────── ─────────────────── ───────── │ │ 01HXR7DF execute errored local/format-codebase local/pkg-workers 00:03:07 │ ╰──────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Filters ───────────────────╮ │ Phase: (any) │ │ State: failed │ │ Project: (any) │ │ Action: (any) │ ╰─────────────────────────────╯
╭─ Summary ─────────────╮ │ Total: 1 │ │ Processing: 0 │ │ Completed: 0 │ │ Errored: 1 │ ╰───────────────────────╯
✓ OK 1 plan listed
Example 11: Complex Graph Actor for Multi-Stage Code Review
Scenario: A team builds a custom multi-stage code review actor that performs security analysis, performance analysis, and style checking in a graph topology — with each stage running a different specialized sub-actor.
Complexity: Expert. Custom actor graph, multiple sub-actors, advanced actor configuration.
Step 1: Define Specialized Sub-Actors
Create actors/review-pipeline.yaml:
name: local/review-pipeline cleveragents: version: "3.0" template_engine: JINJA2 default_actor: orchestratorglobal_context: review_standards: "Google Python Style Guide" severity_levels: "critical, high, medium, low, info"
actors: orchestrator: type: llm config: actor: anthropic/claude-3.5-sonnet temperature: 0.1 system_prompt: | You are a code review orchestrator. You coordinate specialized reviewers and synthesize their findings into a unified review report. Standards: {{ context.review_standards }} Severity levels: {{ context.severity_levels }} memory_enabled: true
security_reviewer: type: llm config: actor: anthropic/claude-3.5-sonnet temperature: 0.0 system_prompt: | You are a security-focused code reviewer. Analyze code for: - Injection vulnerabilities (SQL, command, XSS) - Authentication and authorization flaws - Sensitive data exposure - Insecure cryptographic practices - Dependency vulnerabilities Report findings with severity levels: {{ context.severity_levels }}
performance_reviewer: type: llm config: actor: anthropic/claude-3.5-sonnet temperature: 0.0 system_prompt: | You are a performance-focused code reviewer. Analyze code for: - N+1 query patterns - Unbounded loops or recursion - Memory leaks and excessive allocation - Missing caching opportunities - Blocking operations in async contexts Report findings with severity levels: {{ context.severity_levels }}
style_reviewer: type: llm config: actor: anthropic/claude-3.5-sonnet temperature: 0.0 system_prompt: | You are a code style and maintainability reviewer. Standards: {{ context.review_standards }} Analyze code for: - Naming conventions - Function/class design - Documentation completeness - Test coverage patterns - Code duplication
routes: review_graph: type: graph entry_point: dispatch checkpointing: true nodes: dispatch: type: agent agent: orchestrator security: type: agent agent: security_reviewer performance: type: agent agent: performance_reviewer style: type: agent agent: style_reviewer synthesize: type: agent agent: orchestrator edges: - from: dispatch to: security - from: dispatch to: performance - from: dispatch to: style - from: security to: synthesize - from: performance to: synthesize - from: style to: synthesize parallel_execution: true
$ agents actor add --config ./actors/review-pipeline.yaml \ --skill local/code-analysis \ --skill local/file-ops╭─ Actor Added ──────────────────────────────────╮ │ Name: local/review-pipeline │ │ Provider: anthropic │ │ Model: claude-3.5-sonnet │ │ Default: yes │ │ Unsafe: no │ │ Type: graph │ ╰────────────────────────────────────────────────╯
╭─ Config ─────────────────────────────────────────────╮ │ Path: ./actors/review-pipeline.yaml │ │ Hash: 4f7a1e3 │ │ Options: 6 │ │ Nodes: 5 │ │ Edges: 6 │ ╰──────────────────────────────────────────────────────╯
╭─ Graph ────────────────────────────────────────────────────╮ │ Route: review_graph │ │ Entry Point: dispatch │ │ Checkpointing: yes │ │ Parallel Execution: yes │ │ │ │ Node Type Agent │ │ ─────────── ───── ────────────────────────────── │ │ dispatch agent orchestrator │ │ security agent security_reviewer │ │ performance agent performance_reviewer │ │ style agent style_reviewer │ │ synthesize agent orchestrator │ │ │ │ From To │ │ ─────────── ─────────── │ │ dispatch security │ │ dispatch performance │ │ dispatch style │ │ security synthesize │ │ performance synthesize │ │ style synthesize │ ╰────────────────────────────────────────────────────────────╯
╭─ Capabilities ──────────────────────╮ │ - multi-stage code review │ │ - security analysis │ │ - performance analysis │ │ - style and maintainability checks │ │ - unified report synthesis │ ╰─────────────────────────────────────╯
╭─ Skills ──────────────────────────────────────────╮ │ Skill Status │ │ ─────────────────── ──────────────────────── │ │ local/code-analysis attached │ │ local/file-ops attached │ ╰───────────────────────────────────────────────────╯
✓ OK Actor added
Step 2: Create the Review Action Using the Graph Actor
Create actions/deep-review.yaml:
name: local/deep-review
description: "Multi-stage deep code review with specialized reviewers"
definition_of_done: |
- Security analysis complete with all findings classified by severity
- Performance analysis complete with all findings classified by severity
- Style analysis complete with all findings classified by severity
- Unified review report generated with prioritized action items
- Critical and high-severity issues have suggested fixes
strategy_actor: local/review-pipeline
execution_actor: local/review-pipeline
read_only: true
reusable: true
state: available
args:
- name: target_paths
type: string
required: true
description: "Comma-separated file or directory paths to review"
- name: review_depth
type: string
required: false
default: "standard"
description: "Review depth: quick, standard, thorough"
$ agents action create --config ./actions/deep-review.yaml╭─ Action Created ─────────────────────────────────────────────╮ │ Name: local/deep-review │ │ State: available │ │ Strategy Actor: local/review-pipeline │ │ Execution Actor: local/review-pipeline │ │ Reusable: yes │ │ Read Only: yes │ │ Created: 2026-02-11 16:30 │ ╰──────────────────────────────────────────────────────────────╯
╭─ Definition of Done ──────────────────────────────────────────────────────╮ │ - Security analysis complete with all findings classified by severity │ │ - Performance analysis complete with all findings classified by severity │ │ - Style analysis complete with all findings classified by severity │ │ - Unified review report generated with prioritized action items │ │ - Critical and high-severity issues have suggested fixes │ ╰───────────────────────────────────────────────────────────────────────────╯
╭─ Arguments ───────────────────────────────────────────────────────────────────────╮ │ Name Type Required Description │ │ ──────────── ────── ──────── ────────────────────────────────────────────── │ │ target_paths string yes Comma-separated file or directory paths to review │ │ review_depth string no Review depth: quick, standard, thorough │ ╰───────────────────────────────────────────────────────────────────────────────────╯
╭─ Automation ──────────────────────────╮ │ Profile: supervised │ │ Source: default │ ╰───────────────────────────────────────╯
╭─ Usage ────────────────────────────────────────────────────────────────────────╮ │ agents plan use local/deep-review local/<project> │ │ --arg target_paths="src/auth/,src/routes/" │ │ --arg review_depth="thorough" (optional, default: "standard") │ ╰────────────────────────────────────────────────────────────────────────────────╯
✓ OK Action created
$ agents plan use \ --automation-profile trusted \ --arg target_paths="src/auth/,src/routes/" \ --arg review_depth="thorough" \ local/deep-review local/api-service╭─ Plan Created ───────────────────────────────────────╮ │ Plan: 01HXR8D2E3F4G5H6J7K8L9M0N1 │ │ Phase: strategize │ │ State: processing │ │ Action: local/deep-review │ │ Project: local/api-service │ │ Automation: trusted │ │ Attempt: 1 │ ╰──────────────────────────────────────────────────────╯
╭─ Inputs ───────────────────────────────────────╮ │ - target_paths="src/auth/,src/routes/" │ │ - review_depth="thorough" │ │ - automation_profile=trusted │ ╰────────────────────────────────────────────────╯
╭─ Actors ────────────────────────────────────────╮ │ Strategy: local/review-pipeline │ │ Execution: local/review-pipeline │ │ Estimation: (none) │ │ Actor Type: graph (5 nodes, 6 edges) │ ╰─────────────────────────────────────────────────╯
╭─ Automation ───────────────────────────╮ │ Profile: trusted │ │ Source: plan override │ │ Read-Only: yes │ ╰────────────────────────────────────────╯
╭─ Context ──────────────────────────╮ │ Resources: 1 (git-checkout) │ │ Indexed Files: 312 │ │ View: strategize │ │ Hot Token Budget: 24,000 │ ╰────────────────────────────────────╯
╭─ Next Steps ─────────────────────────────────────────────────────────╮ │ Trusted profile — strategize and execute will proceed automatically. │ │ - agents plan status 01HXR8D2E3F4G5H6J7K8L9M0N1 │ │ - agents plan tree 01HXR8D2E3F4G5H6J7K8L9M0N1 │ ╰──────────────────────────────────────────────────────────────────────╯
✓ OK Plan created
The graph actor executes review_graph with parallel fan-out. The dispatch node (orchestrator) analyzes the target paths and prepares review context, then three edges fire concurrently: security, performance, and style nodes each run their specialized sub-actor in parallel. Once all three complete, their outputs converge at the synthesize node, where the orchestrator merges findings into a single prioritized report. Because read_only: true, no file modifications are produced — only the review report.
Example 12: Large-Scale Feature Implementation with Hierarchical Decomposition
Scenario: A startup is building a new notification system that spans the backend API, a message queue, a worker service, and a frontend dashboard. The feature requires changes across 4 projects and should decompose into manageable child plans.
Complexity: Expert. Multi-project, hierarchical child plans, deep decision tree, invariant reconciliation, full plan lifecycle.
Step 1: Set Up All Projects
# Register the API repo resource (full output shown)$ agents resource add git-checkout local/api-repo --path /repos/api --branch main
╭─ Resource ──────────────────────────────────────╮ │ Name: local/api-repo │ │ ID: 01HXRA1A2B3C4D5E6F7G8H9J0K │ │ Type: git-checkout │ │ Physical/Virtual: physical │ │ Path: /repos/api │ │ Branch: main │ │ Created: 2026-02-11 14:00 │ ╰─────────────────────────────────────────────────╯
╭─ Auto-discovered Children ─────────────────────────────────────────╮ │ ID Type Status │ │ ─────────────── ────────────── ───────────────── │ │ 01HXRA1A2B3C… git created │ │ 01HXRA1A2B3D… git-remote created │ │ 01HXRA1A2B3E… git-branch created │ │ + 48 git-commit resources │ │ + 312 git-tree-entry resources │ │ + 6 fs-directory + 42 fs-file │ ╰────────────────────────────────────────────────────────────────────╯
╭─ Capabilities ─────────────────╮ │ Readable: yes │ │ Writable: yes │ │ Sandboxable: yes │ │ Checkpointable: yes │ │ Sandbox Strategy: git_worktree │ ╰────────────────────────────────╯
✓ OK Resource registered (412 child resources discovered)
# Register the remaining 3 resources (condensed output)
$ agents resource add git-checkout local/worker-repo --path /repos/worker --branch main ✓ OK Resource registered — Name: local/worker-repo (278 children)
$ agents resource add git-checkout local/frontend-repo --path /repos/frontend --branch main ✓ OK Resource registered — Name: local/frontend-repo (524 children)
$ agents resource add git-checkout local/protos-repo --path /repos/protos --branch main ✓ OK Resource registered — Name: local/protos-repo (96 children)
# Create the Backend API project (full output shown)$ agents project create -d "Backend API"
--resource local/api-repo
--resource local/protos-repo
--invariant "All new endpoints must have OpenAPI docs"
--invariant "All database changes require migration scripts"
local/api╭─ Project Created ─────────────────────────╮ │ Name: local/api │ │ Description: Backend API │ │ Type: local │ │ Created: 2026-02-11 14:01 │ ╰───────────────────────────────────────────╯
╭─ Linked Resources ───────────────────────────────────────╮ │ Resource Type Read-Only │ │ ───────────────── ────────────── ───────── │ │ local/api-repo git-checkout no │ │ local/protos-repo git-checkout no │ ╰──────────────────────────────────────────────────────────╯
╭─ Invariants ─────────────────────────────────────────────╮ │ 1. All new endpoints must have OpenAPI docs │ │ 2. All database changes require migration scripts │ ╰──────────────────────────────────────────────────────────╯
╭─ Defaults ──────────────────────────────╮ │ Sandbox: git_worktree │ │ Validations: 0 │ │ Context Filters: none │ │ Automation Profile: (inherits global) │ ╰─────────────────────────────────────────╯
✓ OK Project created
# Create the remaining 3 projects (condensed output)
$ agents project create -d "Background worker service"
--resource local/worker-repo
--resource local/protos-repo
--invariant "Workers must be idempotent"
--invariant "All queue consumers must implement dead-letter handling"
local/worker ✓ OK Project created — Name: local/worker$ agents project create -d "Frontend dashboard"
--resource local/frontend-repo
--invariant "All components must have Storybook stories"
--invariant "Accessibility: WCAG 2.1 AA compliance required"
local/frontend ✓ OK Project created — Name: local/frontend
$ agents project create -d "Shared protobuf definitions"
--resource local/protos-repo
--invariant "Proto changes must be backward-compatible (no field renumbering)"
local/protos ✓ OK Project created — Name: local/protos
# Register validations — first one with full output$ agents validation add --config validations/api-tests.yaml --required local/api-tests
╭─ Validation Registered ─────────────────────────────╮ │ Name: local/api-tests │ │ Command: pytest tests/ && mypy src/ │ │ Mode: required │ │ Timeout: 300s │ ╰─────────────────────────────────────────────────────╯
✓ OK Validation registered
$ agents validation add --config validations/worker-tests.yaml --required local/worker-tests ✓ OK Validation registered — Name: local/worker-tests
$ agents validation add --config validations/frontend-tests.yaml --required local/frontend-tests ✓ OK Validation registered — Name: local/frontend-tests
$ agents validation add --config validations/proto-lint.yaml --required local/proto-lint ✓ OK Validation registered — Name: local/proto-lint
# Attach validations to their respective projects
$ agents validation attach --project local/api local/api-tests ✓ OK Validation attached to project local/api
$ agents validation attach --project local/worker local/worker-tests ✓ OK Validation attached to project local/worker
$ agents validation attach --project local/frontend local/frontend-tests ✓ OK Validation attached to project local/frontend
$ agents validation attach --project local/protos local/proto-lint ✓ OK Validation attached to project local/protos
# Global invariant for this effort$ agents invariant add --global
"All inter-service communication must use the shared proto definitions"╭─ Invariant Added ──────────────────────────────────────────────────────────────╮ │ Invariant: All inter-service communication must use the shared proto defs. │ │ Scope: global │ │ ID: inv_01HXRI1A │ ╰────────────────────────────────────────────────────────────────────────────────╯
✓ OK Invariant added
Step 2: Create the Feature Action
Create actions/build-notification-system.yaml:
name: local/build-notification-system
description: "Build a complete notification system spanning multiple services"
long_description: |
Implement a full notification system with:
- Backend API: notification preferences, send endpoints, delivery tracking
- Message Queue: notification event schemas, routing rules
- Worker: email/SMS/push delivery workers with retry logic
- Frontend: notification center UI, preference settings, real-time updates
The system must be designed for reliability (at-least-once delivery),
scalability (async processing via queue), and user control (per-channel
preferences with quiet hours).
definition_of_done: |
- Proto definitions for notification events are complete and compatible
- API endpoints: CRUD for preferences, send notification, list history
- Worker processes: email, SMS, push delivery with retry and dead-letter
- Frontend: notification center, preference settings, real-time badge
- Integration tests verify end-to-end notification flow
- All per-project validations pass
- API docs updated, Storybook stories added strategy_actor: local/refactoring-strategist execution_actor: anthropic/claude-3.5-sonnet estimation_actor: anthropic/claude-3.5-sonnet invariant_actor: local/invariant-resolver automation_profile: cautious reusable: false state: available args:
- name: notification_channels type: string required: true description: "Comma-separated: email, sms, push, in-app"
- name: priority_levels type: string required: false default: "critical,high,normal,low"
- name: max_retry_attempts type: integer required: false default: 5 invariants:
- "Proto definitions must be implemented and reviewed before any service code"
- "Each service must be deployable independently after its changes"
- "Frontend must gracefully degrade if the notification API is unavailable"
"All notification delivery must be async — API endpoints return immediately"
$ agents action create --config ./actions/build-notification-system.yaml╭─ Action Created ──────────────────────────────────────────────────╮ │ Name: local/build-notification-system │ │ ID: 01HXRB1A2B3C4D5E6F7G8H9J0 │ │ State: available │ │ Strategy Actor: local/refactoring-strategist │ │ Execution Actor: anthropic/claude-3.5-sonnet │ │ Estimation Actor: anthropic/claude-3.5-sonnet │ │ Invariant Actor: local/invariant-resolver │ │ Reusable: no │ │ Read Only: no │ │ Created: 2026-02-11 14:02 │ ╰───────────────────────────────────────────────────────────────────╯
╭─ Definition of Done ───────────────────────────────────────────────────╮ │ - Proto definitions for notification events are complete/compatible │ │ - API endpoints: CRUD for preferences, send notification, list hist. │ │ - Worker: email, SMS, push delivery with retry and dead-letter │ │ - Frontend: notification center, preference settings, real-time badge │ │ - Integration tests verify end-to-end notification flow │ │ - All per-project validations pass │ │ - API docs updated, Storybook stories added │ ╰────────────────────────────────────────────────────────────────────────╯
╭─ Arguments ──────────────────────────────────────────────────────────────────────────╮ │ Name Type Required Description │ │ ────────────────────── ──────── ──────── ────────────────────────────────── │ │ notification_channels string yes Comma-separated: email, sms, push │ │ priority_levels string no (default: critical,high,normal,low) │ │ max_retry_attempts integer no (default: 5) │ ╰──────────────────────────────────────────────────────────────────────────────────────╯
╭─ Automation ──────────────────────────╮ │ Profile: cautious │ │ Source: action config │ ╰───────────────────────────────────────╯
╭─ Action Invariants ───────────────────────────────────────────────────────────────────╮ │ 1. Proto definitions must be implemented and reviewed before any service code │ │ 2. Each service must be deployable independently after its changes │ │ 3. Frontend must gracefully degrade if the notification API is unavailable │ │ 4. All notification delivery must be async — API endpoints return immediately │ ╰───────────────────────────────────────────────────────────────────────────────────────╯
╭─ Usage ────────────────────────────────────────────────────────────────────────────╮ │ agents plan use local/build-notification-system local/protos │ │ --arg notification_channels="email,push,in-app" │ ╰────────────────────────────────────────────────────────────────────────────────────╯
✓ OK Action created
Step 3: Initiate the Plan
$ agents plan use \ --arg notification_channels="email,push,in-app" \ --arg priority_levels="critical,high,normal,low" \ --arg max_retry_attempts=5 \ local/build-notification-system \ local/protos local/api local/worker local/frontend╭─ Plan Created ────────────────────────────────────────────────╮ │ Plan: 01HXRF1A2B3C4D5E6F7G8H9J0 │ │ Action: local/build-notification-system │ │ Projects: local/protos, local/api, │ │ local/worker, local/frontend │ │ Automation: cautious │ │ Attempt: 1 │ ╰───────────────────────────────────────────────────────────────╯
╭─ Inputs ───────────────────────────────────────────────────────────╮ │ - notification_channels="email,push,in-app" │ │ - priority_levels="critical,high,normal,low" │ │ - max_retry_attempts=5 │ │ - automation_profile=cautious │ ╰────────────────────────────────────────────────────────────────────╯
╭─ Actors ────────────────────────────────────────────╮ │ Strategy: local/refactoring-strategist │ │ Execution: anthropic/claude-3.5-sonnet │ │ Estimation: anthropic/claude-3.5-sonnet │ │ Invariant: local/invariant-resolver │ ╰─────────────────────────────────────────────────────╯
╭─ Automation ──────────────────────────────╮ │ Profile: cautious │ │ Source: action config │ │ Read-Only: no │ ╰───────────────────────────────────────────╯
╭─ Context ──────────────────────────╮ │ Resources: 4 (git-checkout) │ │ Indexed Files: 1,310 │ │ View: strategize │ │ Hot Token Budget: 32,000 │ ╰────────────────────────────────────╯
╭─ Next Steps ──────────────────────────────────────────────╮ │ - agents plan tree 01HXRF1A2B3C4D5E6F7G8H9J0 │ │ - agents plan status 01HXRF1A2B3C4D5E6F7G8H9J0 │ │ - agents plan execute 01HXRF1A2B3C4D5E6F7G8H9J0 │ ╰───────────────────────────────────────────────────────────╯
✓ OK Plan created
The plan (01HXRF1A2B3C4D5E6F7G8H9J0) enters Strategize. The strategy actor:
- Reconciles invariants from all four projects plus global plus action-level
- Decomposes the work into ordered child plans:
- Phase 1: Proto definitions (local/protos)
- Phase 2 (parallel): API endpoints (local/api) + Worker skeleton (local/worker)
- Phase 3 (parallel): Worker delivery logic (local/worker) + Frontend notification center (local/frontend)
- Phase 4: Integration wiring and end-to-end tests
Because the cautious profile uses confidence thresholds, the system pauses on complex architectural decisions (e.g., "Should notifications use a fan-out or point-to-point queue topology?") but proceeds automatically on routine decisions.
Step 4: Monitor and Guide the Hierarchical Execution
# View the full plan tree with child plans$ agents plan tree 01HXRF1A2B3C4D5E6F7G8H9J0
╭─ Decision Tree ─────────────────────────────────────────────────────────────────────────────────────────╮ │ Plan: 01HXRF1A2B3C4D5E6F7G8H9J0 │ │ ├─ [prompt_definition] "Build notification system across 4 projects" │ │ ├─ [invariant_enforced] "All inter-service communication must use shared proto defs" │ │ ├─ [invariant_enforced] "Proto definitions must be implemented before any service code" │ │ ├─ [invariant_enforced] "Each service must be deployable independently after its changes" │ │ ├─ [invariant_enforced] "Frontend must gracefully degrade if notification API unavailable" │ │ ├─ [invariant_enforced] "All notification delivery must be async" │ │ ├─ [strategy_choice] "Phase 1: proto definitions first (blocking)" (conf: 0.95) │ │ ├─ [strategy_choice] "Phase 2: API + worker skeleton in parallel" (conf: 0.88) │ │ ├─ [strategy_choice] "Phase 3: worker delivery + frontend UI in parallel" (conf: 0.84) │ │ ├─ [strategy_choice] "Phase 4: integration wiring and e2e tests" (conf: 0.90) │ │ ├─ [strategy_choice] "Queue topology: fan-out vs point-to-point?" (conf: 0.55 — PAUSED) │ │ ├─ [child_plan] 01HXRC1A2B3C4D5E6F7G8H9J0 (Phase 1: proto defs) [execute/complete] │ │ ├─ [child_plan] 01HXRC2B3C4D5E6F7G8H9J0K1 (Phase 2a: API endpoints) [execute/processing] │ │ ├─ [child_plan] 01HXRC3C4D5E6F7G8H9J0K1L2 (Phase 2b: worker skeleton) [execute/processing] │ │ ├─ [child_plan] 01HXRC4D5E6F7G8H9J0K1L2M3 (Phase 3a: worker delivery) [strategize/pending] │ │ ├─ [child_plan] 01HXRC5E6F7G8H9J0K1L2M3N4 (Phase 3b: frontend UI) [strategize/pending] │ │ └─ [child_plan] 01HXRC6F7G8H9J0K1L2M3N4P5 (Phase 4: integration) [strategize/pending] │ ╰─────────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Tree Summary ──────────────────╮ │ Nodes: 17 │ │ Depth: 2 │ │ Child Plans: 6 (1 done, 2 exec) │ │ Invariants: 5 │ │ Superseded: 0 (hidden) │ ╰─────────────────────────────────╯
╭─ Decision IDs (for correction) ──────────────────────────╮ │ Root: 01HXRD1A2B3C4D5E6F7G8H9J │ │ Invariant 1: 01HXRD1B3C4D5E6F7G8H9J0 │ │ Invariant 2: 01HXRD1C4D5E6F7G8H9J0K1 │ │ Invariant 3: 01HXRD1D5E6F7G8H9J0K1L2 │ │ Invariant 4: 01HXRD1E6F7G8H9J0K1L2M3 │ │ Invariant 5: 01HXRD1F7G8H9J0K1L2M3N4 │ │ Strategy 1: 01HXRD2A3B4C5D6E7F8G9H0 │ │ Strategy 2: 01HXRD2B4C5D6E7F8G9H0J1 │ │ Strategy 3: 01HXRD2C5D6E7F8G9H0J1K2 │ │ Strategy 4: 01HXRD2D6E7F8G9H0J1K2L3 │ │ Paused: 01HXRD2E7F8G9H0J1K2L3M4 │ ╰──────────────────────────────────────────────────────────╯
✓ OK Decision tree rendered
# The system paused at a decision about queue topology (confidence 0.55 < threshold 0.7)$ agents plan explain 01HXRD2E7F8G9H0J1K2L3M4
╭─ Decision ──────────────────────────────────────────────────╮ │ ID: 01HXRD2E7F8G9H0J1K2L3M4 │ │ Type: strategy_choice │ │ Question: Queue topology for notification delivery │ │ Chosen: (none — paused for user input) │ │ Confidence: 0.55 │ │ Plan: 01HXRF1A2B3C4D5E6F7G8H9J0 │ │ Sequence: 5 of 5 strategy choices │ │ Created: 2026-02-11 14:04 │ ╰─────────────────────────────────────────────────────────────╯
╭─ Alternatives Considered ──────────────────────────────────────────────────────╮ │ 1. Fan-out exchange: one exchange per event type, separate queues per channel │ │ 2. Point-to-point: single queue, worker routes by notification type │ │ 3. Topic-based: single exchange with routing keys for channel + priority │ ╰────────────────────────────────────────────────────────────────────────────────╯
╭─ Impact ─────────────────────────╮ │ Downstream Decisions: 4 │ │ Downstream Child Plans: 3 │ │ Artifacts Produced: 0 (pending) │ │ Correction Impact: high │ ╰──────────────────────────────────╯
╭─ Rationale ──────────────────────────────────────────────────────────╮ │ The confidence is low because both fan-out and topic-based patterns │ │ are viable. Fan-out gives independent scaling per channel but adds │ │ more exchanges. Topic-based is simpler but couples routing logic. │ │ The existing codebase uses RabbitMQ but has no precedent for this │ │ pattern — no strong signal either way. │ ╰──────────────────────────────────────────────────────────────────────╯
╭─ Correction ──────────────────────────────────────────────────╮ │ agents plan correct 01HXRD2E7F8G9H0J1K2L3M4 │ │ --mode revert --guidance "Use fan-out exchange pattern..." │ ╰───────────────────────────────────────────────────────────────╯
✓ OK Decision explained
# Provide guidance to resolve the paused decision$ agents plan prompt 01HXRF1A2B3C4D5E6F7G8H9J0
"Use a fan-out exchange pattern: one exchange per notification event type,
with separate queues per delivery channel. This matches our existing
RabbitMQ setup in the order service."╭─ Guidance Added ────────────────────────────────────────────────────────╮ │ Plan: 01HXRF1A2B3C4D5E6F7G8H9J0 │ │ Guidance: Use a fan-out exchange pattern: one exchange per │ │ notification event type, with separate queues per delivery │ │ channel. This matches our existing RabbitMQ setup. │ │ Scope: next execution step │ │ Phase: strategize │ │ State: awaiting_input → processing │ ╰─────────────────────────────────────────────────────────────────────────╯
╭─ Decision Created ────────────────────────╮ │ Type: user_intervention │ │ ID: 01HXRD3A2B3C4D5E6F7G8H9J │ │ Parent: 01HXRD2E7F8G9H0J1K2L3M4 │ ╰───────────────────────────────────────────╯
╭─ Queue ────╮ │ Pending: 1 │ │ Applied: 0 │ ╰────────────╯
✓ OK Guidance queued
Step 5: Handle a Failure in One Child Plan
The worker delivery child plan fails validation because the SMS provider SDK is not installed. Rather than failing the entire tree:
# Check which child plan failed$ agents plan status 01HXRC4D5E6F7G8H9J0K1L2M3
╭─ Plan Status ──────────────────────────────────────────────╮ │ Plan: 01HXRC4D5E6F7G8H9J0K1L2M3 │ │ Phase: execute │ │ State: errored │ │ Action: local/build-notification-system (child plan) │ │ Project: local/worker │ │ Automation: cautious │ │ Attempt: 1 │ │ Parent: 01HXRF1A2B3C4D5E6F7G8H9J0 │ ╰────────────────────────────────────────────────────────────╯
╭─ Progress ───────────╮ │ ✓ Strategize │ │ ✗ Execute │ │ ○ Apply (skipped) │ ╰──────────────────────╯
╭─ Error Detail ──────────────────────────────────────────────────────────╮ │ Error: Validation failed: ModuleNotFoundError: No module named │ │ 'twilio' — SMS provider SDK not installed │ │ Tool: execute_command (pip install check) │ │ Step: 3 of 8 │ │ Checkpoint: cp_01HXRC4D (sandbox state preserved) │ │ Recoverable: yes — use "agents plan prompt" to provide guidance │ ╰─────────────────────────────────────────────────────────────────────────╯
╭─ Cost ───────────────╮ │ Tokens Used: 14,280 │ │ Cost So Far: $0.048 │ ╰──────────────────────╯
✗ ERROR Plan errored — useagents plan promptto resume oragents plan cancelto abort
# Correct the strategy — tell it to skip SMS for now$ agents plan correct 01HXRD4A2B3C4D5E6F7G8H9J
--mode append
--guidance "Skip SMS delivery for the initial implementation.
Add a placeholder worker that logs a warning.
SMS support will be added in a follow-up action."╭─ Correction ─────────────────────────────────────────────────╮ │ Mode: append │ │ Impact: adds to existing subtree, no rollback │ │ New Decision: 01HXRD5A2B3C4D5E6F7G8H9J │ │ Appended After: 01HXRD4A2B3C4D5E6F7G8H9J │ │ Attempt: 2 │ ╰──────────────────────────────────────────────────────────────╯
╭─ Append Detail ────────────────────────────────────────────────────╮ │ Original decision preserved: yes │ │ Existing artifacts kept: yes │ │ Additional work: re-execute with SMS placeholder │ │ The worker delivery plan will be re-executed. Email and push │ │ workers are kept; SMS worker replaced with a logging stub. │ │ Other child plans (API, frontend, integration) are unaffected. │ ╰────────────────────────────────────────────────────────────────────╯
╭─ Queued ──────────────╮ │ Re-execute: 1 child │ │ ETA: 3m │ ╰───────────────────────╯
✓ OK Append correction queued
Step 6: Apply in Phases
# Apply proto definitions first (other plans depend on this)$ agents plan apply --yes 01HXRC1A2B3C4D5E6F7G8H9J0
╭─ Apply Summary ──────────────────────────────────────────╮ │ Plan: 01HXRC1A2B3C4D5E6F7G8H9J0 │ │ Artifacts: 3 files updated │ │ Changes: 84 insertions, 0 deletions │ │ Project: local/protos │ │ Applied At: 2026-02-11 14:18 │ ╰──────────────────────────────────────────────────────────╯
╭─ Validation ──────────────────────────────────────────╮ │ buf lint: passed │ │ buf breaking: passed (no breaking changes) │ │ Duration: 2.1s │ ╰───────────────────────────────────────────────────────╯
╭─ Sandbox Cleanup ─────────╮ │ Worktree: removed │ │ Branch: merged to main │ │ Checkpoint: archived │ ╰───────────────────────────╯
╭─ Plan Lifecycle ─────────────────────────╮ │ Phase: apply │ │ State: applied │ │ Total Duration: 00:02:41 │ │ Total Cost: $0.024 │ │ Decisions Made: 4 │ │ Child Plans: 0 │ ╰──────────────────────────────────────────╯
✓ OK Changes applied
# Apply API and worker in parallel
$ agents plan apply --yes 01HXRC2B3C4D5E6F7G8H9J0K1 ✓ OK Changes applied — Plan: 01HXRC2B...J0K1 Project: local/api (8 files, +246 -12)
$ agents plan apply --yes 01HXRC3C4D5E6F7G8H9J0K1L2 ✓ OK Changes applied — Plan: 01HXRC3C...K1L2 Project: local/worker (5 files, +178 -4)
# Apply worker delivery and frontend
$ agents plan apply --yes 01HXRC4D5E6F7G8H9J0K1L2M3 ✓ OK Changes applied — Plan: 01HXRC4D...L2M3 Project: local/worker (6 files, +312 -18)
$ agents plan apply --yes 01HXRC5E6F7G8H9J0K1L2M3N4 ✓ OK Changes applied — Plan: 01HXRC5E...M3N4 Project: local/frontend (12 files, +487 -22)
# Apply integration tests
$ agents plan apply --yes 01HXRC6F7G8H9J0K1L2M3N4P5 ✓ OK Changes applied — Plan: 01HXRC6F...N4P5 Projects: local/api, local/worker (4 files, +156 -0)
# Apply the parent plan (finalizes everything)
$ agents plan apply --yes 01HXRF1A2B3C4D5E6F7G8H9J0
╭─ Apply Summary ───────────────────────────────────────────╮ │ Plan: 01HXRF1A2B3C4D5E6F7G8H9J0 │ │ Artifacts: 38 files across 4 projects │ │ Changes: 1,463 insertions, 56 deletions │ │ Projects: local/protos, local/api, │ │ local/worker, local/frontend │ │ Applied At: 2026-02-11 14:32 │ ╰───────────────────────────────────────────────────────────╯
╭─ Validation ─────────────────────────────────────────╮ │ local/protos: passed (buf lint + breaking) │ │ local/api: passed (pytest 48/48, mypy 0 err) │ │ local/worker: passed (pytest 31/31, mypy 0) │ │ local/frontend: passed (tests + lint) │ │ Duration: 34.2s │ ╰──────────────────────────────────────────────────────╯
╭─ Sandbox Cleanup ─────────╮ │ Worktrees: 4 removed │ │ Branches: merged to main │ │ Checkpoints: archived │ ╰───────────────────────────╯
╭─ Plan Lifecycle ────────────────────────╮ │ Phase: apply │ │ State: applied │ │ Total Duration: 00:30:14 │ │ Total Cost: $0.847 │ │ Decisions Made: 42 │ │ Child Plans: 6 (all completed) │ │ Corrections: 1 │ ╰─────────────────────────────────────────╯
╭─ Next Steps ──────╮ │ - Review git diff │ │ - Commit changes │ ╰───────────────────╯
✓ OK Changes applied
Example 13: Custom Automation Profile with Semantic Escalation
Scenario: A team wants fine-grained automation control. They trust the system with routine planning but want human oversight on any decision involving database schema changes or security-sensitive code, regardless of confidence.
Complexity: Intermediate. Custom profile, invariant-driven escalation.
Step 1: Create a Custom Profile
Create profiles/db-cautious.yaml:
name: local/db-cautious description: "Auto for most tasks, manual for database and security decisions"auto_strategize: 0.0 auto_execute: 0.3 auto_apply: 1.0 auto_decisions_strategize: 0.4 auto_decisions_execute: 0.6 auto_validation_fix: 0.5 auto_strategy_revision: 0.8 auto_reversion_from_apply: 0.9 auto_child_plans: 0.3 auto_retry_transient: 0.0 auto_checkpoint_restore: 0.0
require_sandbox: true require_checkpoints: true allow_unsafe_tools: false
$ agents automation-profile add --config ./profiles/db-cautious.yaml╭─ Profile Registered ──────────────────────────────────────────────────────╮ │ Name: local/db-cautious │ │ Description: Auto for most tasks, manual for database and security │ │ decisions │ │ Created: 2026-02-11 09:00 │ ╰───────────────────────────────────────────────────────────────────────────╯
╭─ Confidence Thresholds ────────────────╮ │ auto_strategize: 0.0 │ │ auto_execute: 0.3 │ │ auto_apply: 1.0 │ │ auto_decisions_strategize: 0.4 │ │ auto_decisions_execute: 0.6 │ │ auto_validation_fix: 0.5 │ │ auto_strategy_revision: 0.8 │ │ auto_child_plans: 0.3 │ │ auto_retry_transient: 0.0 │ │ auto_checkpoint_restore: 0.0 │ │ require_sandbox: true │ │ require_checkpoints: true │ │ allow_unsafe_tools: false │ ╰────────────────────────────────────────╯
✓ OK Profile registered
Step 2: Apply to a Project with Database Resources
$ agents config set core.automation-profile local/db-cautious --project local/api-service╭─ Config Updated ─────────────────────────────────╮ │ Key: core.automation-profile │ │ Value: local/db-cautious │ │ Previous: review │ │ Source: config │ │ Scope: project (local/api-service) │ ╰──────────────────────────────────────────────────╯
╭─ Effective ────────────────────────────────────╮ │ Sessions: new sessions on local/api-service │ │ Plans: future plans (unless overridden) │ │ Existing: unchanged │ ╰────────────────────────────────────────────────╯
╭─ Saved To ──────────────────────────╮ │ File: ~/.cleveragents/config.toml │ │ Line: 24 │ ╰─────────────────────────────────────╯
✓ OK Config updated
# Add invariants that force escalation on sensitive operations $ agents invariant add --project local/api-service \ "Any change to database migration files requires explicit human approval"╭─ Invariant Added ───────────────────────────────────────────────────────────────╮ │ Project: local/api-service │ │ Invariant: Any change to database migration files requires explicit human │ │ approval │ │ Scope: project │ │ ID: inv_01HXRBA1A │ ╰─────────────────────────────────────────────────────────────────────────────────╯
✓ OK Invariant added
$ agents invariant add --project local/api-service
"Changes to authentication or authorization logic require security review"╭─ Invariant Added ───────────────────────────────────────────────────────────────╮ │ Project: local/api-service │ │ Invariant: Changes to authentication or authorization logic require │ │ security review │ │ Scope: project │ │ ID: inv_01HXRBA2B │ ╰─────────────────────────────────────────────────────────────────────────────────╯
✓ OK Invariant added
Step 3: Run a Plan and Observe Escalation
$ agents plan use \ --arg target_module="src/auth" \ local/refactor-to-orm local/api-service╭─ Plan Created ─────────────────────────────────────╮ │ Plan ID: 01HXRBG1H2I3J4K5L6M7N8O9P0 │ │ Phase: strategize (running) │ │ Action: local/refactor-to-orm │ │ Project: local/api-service │ │ Automation: local/db-cautious │ │ Attempt: 1 │ ╰────────────────────────────────────────────────────╯
╭─ Inputs ─────────────────────╮ │ - target_module=src/auth │ ╰──────────────────────────────╯
╭─ Actors ───────────────────────────────╮ │ Strategy: anthropic/claude-3.5-sonnet │ │ Execution: anthropic/claude-3.5-sonnet │ │ Invariant: local/invariant-resolver │ ╰────────────────────────────────────────╯
╭─ Plan Invariants ──────────────────────────────────────────────────────────────╮ │ Scope Source Invariant │ │ ──────── ─────── ────────────────────────────────────────────────────── │ │ project config Any change to database migration files requires explicit │ │ human approval │ │ project config Changes to authentication or authorization logic require │ │ security review │ ╰────────────────────────────────────────────────────────────────────────────────╯
╭─ Next Steps ─────────────────────────────────────────╮ │ - agents plan status 01HXRBG1H2I3J4K5L6M7N8O9P0 │ │ - agents plan tree 01HXRBG1H2I3J4K5L6M7N8O9P0 │ ╰──────────────────────────────────────────────────────╯
✓ OK Plan created — strategize in progress
During execution, the system encounters a decision to create a new Alembic migration. The confidence score is 0.82 (the system is fairly confident), but the invariant "database migration files require explicit human approval" causes the Invariant Reconciliation Actor to enforce escalation regardless. The system pauses:
# The plan pauses — check why $ agents plan status 01HXRBG1H2I3J4K5L6M7N8O9P0╭─ Plan Status ───────────────────────────────────────╮ │ Plan: 01HXRBG1H2I3J4K5L6M7N8O9P0 │ │ Phase: execute │ │ State: awaiting_input │ │ Action: local/refactor-to-orm │ │ Project: local/api-service │ │ Automation: local/db-cautious │ │ Attempt: 1 │ ╰─────────────────────────────────────────────────────╯
╭─ Progress ────────────────╮ │ ✓ Strategize │ │ ⏸ Execute (paused) │ │ ○ Apply (waiting) │ ╰───────────────────────────╯
╭─ Escalation ───────────────────────────────────────────────────────────────╮ │ ⚠ Decision paused — invariant forced escalation │ │ Decision: 01HXRBD1E2F3G4H5I6J7K8L9 │ │ Type: implementation_choice │ │ Question: Create Alembic migration for users table schema change? │ │ Confidence: 0.82 (would auto-proceed, but invariant overrides) │ │ Invariant: "Any change to database migration files requires explicit │ │ human approval" │ ╰────────────────────────────────────────────────────────────────────────────╯
╭─ Timing ───────────╮ │ Started: 09:05:12 │ │ Elapsed: 00:02:48 │ │ ETA: (paused) │ ╰────────────────────╯
╭─ Cost ───────────────╮ │ Tokens Used: 18,740 │ │ Cost So Far: $0.062 │ │ Estimated: $0.095 │ ╰──────────────────────╯
⚠ WARN Plan paused — awaiting human input for escalated decision
# Review the decision requiring approval $ agents plan explain 01HXRBD1E2F3G4H5I6J7K8L9╭─ Decision ───────────────────────────────────────────────────────────╮ │ ID: 01HXRBD1E2F3G4H5I6J7K8L9 │ │ Type: implementation_choice │ │ Question: Create Alembic migration for users table schema change? │ │ Chosen: Add nullable
orm_versioncolumn with migration │ │ Confidence: 0.82 │ │ Plan: 01HXRBG1H2I3J4K5L6M7N8O9P0 │ │ Sequence: 4 of 7 │ │ Created: 2026-02-11 09:07 │ ╰──────────────────────────────────────────────────────────────────────╯╭─ Alternatives Considered ────────────────────────────────────────────╮ │ 1. Add nullable
orm_versioncolumn with migration (chosen) │ │ 2. Use a shadow table approach (no schema change) │ │ 3. In-place column rename via ALTER (risky in production) │ ╰──────────────────────────────────────────────────────────────────────╯╭─ Escalation Reason ──────────────────────────────────────────────────╮ │ Invariant Override: inv_01HXRBA1A │ │ Text: "Any change to database migration files requires explicit │ │ human approval" │ │ Effect: Confidence 0.82 would normally auto-proceed (threshold │ │ 0.6), but invariant forces manual escalation regardless │ ╰──────────────────────────────────────────────────────────────────────╯
╭─ Impact ─────────────────────────╮ │ Downstream Decisions: 3 │ │ Downstream Child Plans: 0 │ │ Artifacts Produced: 2 │ │ Correction Impact: medium │ ╰──────────────────────────────────╯
╭─ Rationale ──────────────────────────────────────────────────────────╮ │ The auth module currently uses raw SQL queries. Refactoring to ORM │ │ requires an
orm_versiontracking column on the users table. │ │ Adding as nullable first avoids downtime; the backfill + NOT NULL │ │ constraint can follow in a second migration after data is populated. │ ╰──────────────────────────────────────────────────────────────────────╯╭─ Correction ────────────────────────────────────────────────╮ │ agents plan prompt 01HXRBG1H2I3J4K5L6M7N8O9P0 │ │ "Approved. <guidance>" │ │ agents plan correct 01HXRBD1E2F3G4H5I6J7K8L9 │ │ --mode revert --guidance "Use approach 2 instead..." │ ╰─────────────────────────────────────────────────────────────╯
✓ OK Decision explained
# Approve the migration approach with specific guidance $ agents plan prompt 01HXRBG1H2I3J4K5L6M7N8O9P0 \ "Approved. Use a two-phase migration: first add the column as nullable, \ then in a separate migration add the NOT NULL constraint after backfill."╭─ Guidance Added ──────────────────────────────────────────────────────────╮ │ Plan: 01HXRBG1H2I3J4K5L6M7N8O9P0 │ │ Guidance: Approved. Use a two-phase migration: first add the column as │ │ nullable, then in a separate migration add the NOT NULL │ │ constraint after backfill. │ │ Scope: next execution step │ │ Phase: execute │ │ State: awaiting_input → processing │ ╰───────────────────────────────────────────────────────────────────────────╯
╭─ Decision Created ──────────────────────────────╮ │ Type: user_intervention │ │ ID: 01HXRBE2F3G4H5I6J7K8L9M0 │ │ Parent: 01HXRBD1E2F3G4H5I6J7K8L9 │ ╰─────────────────────────────────────────────────╯
╭─ Queue ────╮ │ Pending: 1 │ │ Applied: 0 │ ╰────────────╯
✓ OK Guidance queued — execution resuming
Example 14: Server Mode — Team Collaboration
Scenario: A distributed team uses CleverAgents in server mode. Multiple engineers share actions, actors, and projects while working on the same codebase from different machines.
Complexity: Intermediate. Server mode, multi-user, namespace management.
Step 1: Connect to the Server
# Configure server connection $ agents config set server.url https://agents.example.com╭─ Config Updated ──────────────────────────────╮ │ Key: server.url │ │ Value: https://agents.example.com │ │ Previous: (not set) │ │ Source: config │ │ Scope: global │ ╰───────────────────────────────────────────────╯
╭─ Saved To ──────────────────────────╮ │ File: ~/.cleveragents/config.toml │ │ Line: 3 │ ╰─────────────────────────────────────╯
✓ OK Config updated
$ agents config set server.token "tok_01HXR..."
╭─ Config Updated ──────────────────────────╮ │ Key: server.token │ │ Value: tok_01HXR... (redacted) │ │ Previous: (not set) │ │ Source: config │ │ Scope: global │ ╰───────────────────────────────────────────╯
╭─ Saved To ──────────────────────────╮ │ File: ~/.cleveragents/config.toml │ │ Line: 4 │ ╰─────────────────────────────────────╯
✓ OK Config updated
$ agents config set core.namespace myteam
╭─ Config Updated ──────────────────────╮ │ Key: core.namespace │ │ Value: myteam │ │ Previous: local │ │ Source: config │ │ Scope: global │ ╰───────────────────────────────────────╯
╭─ Saved To ──────────────────────────╮ │ File: ~/.cleveragents/config.toml │ │ Line: 5 │ ╰─────────────────────────────────────╯
✓ OK Config updated
# Verify connection $ agents diagnostics╭─ Checks ────────────────────────────────────────────────╮ │ Check Status Details │ │ ─────────────── ────── ─────────────────────── │ │ Config file OK readable │ │ Database OK writable │ │ Anthropic key OK configured │ │ OPENAI_API_KEY WARN missing │ │ Disk space OK 4.8 GB free │ │ Text index OK tantivy 0.22 │ │ Vector index OK faiss (CPU) │ │ Graph store OK neo4j 5.15 │ │ File permissions OK data dir r/w │ │ Git OK git 2.43.0 │ │ Server OK connected (agents.example.com) │ │ Namespace OK myteam (4 members) │ ╰─────────────────────────────────────────────────────────╯
╭─ Summary ─────────╮ │ Checks: 12 total │ │ Warnings: 1 │ │ Errors: 0 │ │ Duration: 0.8s │ ╰───────────────────╯
╭─ Server ────────────────────────────╮ │ URL: https://agents.example.com │ │ Auth: valid (tok_01HXR...) │ │ Namespace: myteam │ │ Members: 4 │ │ Shared Actions: 7 │ │ Shared Actors: 3 │ │ Latency: 42ms │ ╰─────────────────────────────────────╯
╭─ Recommendations ──────────────────────────────╮ │ - Set OPENAI_API_KEY to enable OpenAI models │ ╰────────────────────────────────────────────────╯
⚠ WARN 1 warning requires attention
Step 2: Publish Shared Resources
# Publish an action to the team namespace (visible to all team members) $ agents action create --config ./actions/generate-tests.yaml╭─ Action Created ──────────────────────────────────────╮ │ Name: myteam/generate-tests │ │ ID: 01HXRC1A2B3C4D5E6F7G8H9J0K │ │ State: available │ │ Strategy Actor: anthropic/claude-3.5-sonnet │ │ Execution Actor: anthropic/claude-3.5-sonnet │ │ Reusable: yes │ │ Read Only: no │ │ Created: 2026-02-11 09:15 │ ╰───────────────────────────────────────────────────────╯
╭─ Definition of Done ───────────────────────────────────╮ │ All target modules have test coverage above 80% │ │ Tests pass in CI without flaky failures │ ╰────────────────────────────────────────────────────────╯
╭─ Arguments ─────────────────────────────────────────────────────╮ │ Name Type Required Description │ │ ────────────── ────── ──────── ─────────────────────── │ │ target_module string yes Module path to generate tests │ ╰─────────────────────────────────────────────────────────────────╯
╭─ Automation ────────────────────────╮ │ Profile: supervised │ │ Source: action default │ ╰─────────────────────────────────────╯
╭─ Usage ───────────────────────────────────────────────────────────╮ │ agents plan use myteam/generate-tests local/my-project │ │ --arg target_module="src/payments" │ ╰───────────────────────────────────────────────────────────────────╯
✓ OK Action created
# Publish a shared actor $ agents actor add --config ./actors/review-pipeline.yaml╭─ Actor Added ──────────────────────────────────╮ │ Name: myteam/review-pipeline │ │ Provider: anthropic │ │ Model: claude-3.5-sonnet │ │ Default: no │ │ Unsafe: no │ │ Type: graph │ ╰────────────────────────────────────────────────╯
╭─ Config ───────────────────────────────╮ │ Path: ./actors/review-pipeline.yaml │ │ Hash: 5a7c2e1 │ │ Options: 3 │ │ Nodes: 5 │ │ Edges: 6 │ ╰────────────────────────────────────────╯
╭─ Capabilities ──────────────────╮ │ - multi-stage code review │ │ - security analysis │ │ - performance analysis │ │ - style checking │ ╰─────────────────────────────────╯
╭─ Tools ────────────────────────╮ │ Tool Read-Only Safe │ │ ──────────── ───────── ──── │ │ read_file yes yes │ │ search_files yes yes │ │ git_diff yes yes │ │ run_tests yes yes │ ╰────────────────────────────────╯
✓ OK Actor added
# List available actions (team namespace) $ agents action list --namespace myteam╭─ Actions ────────────────────────────────────────────────────────────────────────────────────────────────╮ │ Name State Strategy Actor Execution Actor Reusable │ │ ──────────────────────── ───────── ─────────────────────────── ─────────────────────────── ──────── │ │ myteam/generate-tests available anthropic/claude-3.5-sonnet anthropic/claude-3.5-sonnet ✓ │ │ myteam/deep-review available myteam/review-pipeline myteam/review-pipeline ✓ │ │ myteam/format-codebase available anthropic/claude-3.5-sonnet anthropic/claude-3.5-sonnet ✓ │ │ myteam/security-audit available anthropic/claude-3.5-sonnet anthropic/claude-3.5-sonnet ✓ │ │ myteam/refactor-to-orm available anthropic/claude-3.5-sonnet anthropic/claude-3.5-sonnet ✗ │ ╰──────────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Filters ──────────────────╮ │ State: (any) │ │ Namespace: myteam │ ╰────────────────────────────╯
╭─ Summary ──────────────────╮ │ Total: 5 │ │ Available: 5 │ │ Archived: 0 │ ╰────────────────────────────╯
✓ OK 5 actions listed
Step 3: Use Shared Resources from Another Machine
On a different developer's machine:
# Configure server connection $ agents config set server.url https://agents.example.com╭─ Config Updated ──────────────────────────────╮ │ Key: server.url │ │ Value: https://agents.example.com │ │ Previous: (not set) │ │ Source: config │ │ Scope: global │ ╰───────────────────────────────────────────────╯
✓ OK Config updated
$ agents config set server.token "tok_01HXS..."
╭─ Config Updated ──────────────────────────╮ │ Key: server.token │ │ Value: tok_01HXS... (redacted) │ │ Previous: (not set) │ │ Source: config │ │ Scope: global │ ╰───────────────────────────────────────────╯
✓ OK Config updated
$ agents config set core.namespace myteam
╭─ Config Updated ──────────────────────╮ │ Key: core.namespace │ │ Value: myteam │ │ Previous: local │ │ Source: config │ │ Scope: global │ ╰───────────────────────────────────────╯
✓ OK Config updated
# Use the team's shared action $ agents plan use \ --arg target_module="src/payments" \ myteam/generate-tests local/payment-service╭─ Plan Created ─────────────────────────────────────────╮ │ Plan ID: 01HXRC2B3C4D5E6F7G8H9J0K1L2 │ │ Phase: strategize (running) │ │ Action: myteam/generate-tests │ │ Project: local/payment-service │ │ Automation: supervised │ │ Attempt: 1 │ ╰────────────────────────────────────────────────────────╯
╭─ Inputs ──────────────────────╮ │ - target_module=src/payments │ ╰───────────────────────────────╯
╭─ Actors ──────────────────────────────────╮ │ Strategy: anthropic/claude-3.5-sonnet │ │ Execution: anthropic/claude-3.5-sonnet │ │ Source: server (myteam) │ ╰───────────────────────────────────────────╯
╭─ Context ───────────────────────╮ │ Resources: 1 (repo) │ │ Indexed Files: 189 │ │ View: strategize │ │ Hot Token Budget: 12,000 │ ╰─────────────────────────────────╯
╭─ Next Steps ──────────────────────────────────────────────╮ │ - agents plan status 01HXRC2B3C4D5E6F7G8H9J0K1L2 │ │ - agents plan tree 01HXRC2B3C4D5E6F7G8H9J0K1L2 │ ╰───────────────────────────────────────────────────────────╯
✓ OK Plan created — strategize in progress
Step 4: Monitor Across the Team
# List all plans across the team $ agents plan list --namespace myteam╭─ Plans ────────────────────────────────────────────────────────────────────────────────────────────────────────╮ │ ID Phase State Action Project User Elapsed │ │ ──────── ──────── ────────── ────────────────────── ───────────────────── ─────── ───────── │ │ 01HXRC2B execute processing myteam/generate-tests local/payment-service jdoe 00:04:12 │ │ 01HXRC3C apply applied myteam/deep-review local/api-service asmith 00:08:47 │ │ 01HXRC4D execute processing myteam/generate-tests local/api-service mwong 00:02:35 │ │ 01HXRC5E apply applied myteam/format-codebase local/frontend blee 00:01:58 │ │ 01HXRC6F execute awaiting myteam/security-audit local/worker-service asmith 00:06:20 │ ╰────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Filters ──────────────────╮ │ Phase: (any) │ │ State: (any) │ │ Namespace: myteam │ │ Project: (any) │ │ Action: (any) │ ╰────────────────────────────╯
╭─ Summary ─────────────╮ │ Total: 5 │ │ Processing: 2 │ │ Completed: 2 │ │ Awaiting: 1 │ │ Errored: 0 │ ╰───────────────────────╯
✓ OK 5 plans listed
# See who is running what — filter to executing state $ agents plan list --namespace myteam --state processing╭─ Plans ────────────────────────────────────────────────────────────────────────────────────────────────────────╮ │ ID Phase State Action Project User Elapsed │ │ ──────── ─────── ────────── ───────────────────── ───────────────────── ─────── ───────── │ │ 01HXRC2B execute processing myteam/generate-tests local/payment-service jdoe 00:04:12 │ │ 01HXRC4D execute processing myteam/generate-tests local/api-service mwong 00:02:35 │ ╰────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Filters ──────────────────────╮ │ Phase: (any) │ │ State: processing │ │ Namespace: myteam │ │ Project: (any) │ │ Action: (any) │ ╰────────────────────────────────╯
╭─ Summary ─────────────╮ │ Total: 2 │ │ Processing: 2 │ │ Completed: 0 │ │ Errored: 0 │ ╰───────────────────────╯
✓ OK 2 plans listed
Example 15: Disaster Recovery — Rollback a Failed Apply
Scenario: A plan was applied, but the changes broke production. The team needs to understand what happened and roll back.
Complexity: Intermediate. Rollback, checkpoint restore, post-mortem analysis.
Step 1: Investigate the Failed Apply
# Check the plan status $ agents plan status 01HXRCF1G2H3I4J5K6L7M8N9O0╭─ Plan Status ───────────────────────────────────────╮ │ Plan: 01HXRCF1G2H3I4J5K6L7M8N9O0 │ │ Phase: apply │ │ State: errored │ │ Action: local/optimize-db-connections │ │ Project: local/api-service │ │ Automation: trusted │ │ Attempt: 1 │ ╰─────────────────────────────────────────────────────╯
╭─ Progress ─────────────────────╮ │ ✓ Strategize │ │ ✓ Execute │ │ ✗ Apply (post-apply failure) │ ╰────────────────────────────────╯
╭─ Timing ──────────────────────────╮ │ Started: 2026-02-11 08:30:00 │ │ Applied At: 2026-02-11 08:42:18 │ │ Failed At: 2026-02-11 08:44:05 │ │ Total Duration: 00:14:05 │ ╰───────────────────────────────────╯
╭─ Result ───────────────────────────────────────────────╮ │ Decisions Made: 6 │ │ Child Plans: 0 │ │ Artifacts: 4 files updated │ │ Validations: 2/3 passed (1 FAILED post-apply) │ │ Total Cost: $0.072 │ ╰────────────────────────────────────────────────────────╯
╭─ Error ──────────────────────────────────────────────────────────────────────╮ │ Post-apply health check failed: connection pool exhaustion detected │ │ Impact: API response times increased 10x, 503 errors in production │ │ Root Cause Decision: 01HXRCD3E4F5G6H7I8J9K0L1 │ ╰──────────────────────────────────────────────────────────────────────────────╯
╭─ Recovery Options ──────────────────────────────────────────────────╮ │ - agents plan rollback — restore to a previous checkpoint │ │ - agents plan correct — revert and re-execute with guidance │ │ - agents plan explain — investigate the root cause decision │ ╰─────────────────────────────────────────────────────────────────────╯
✗ ERROR Plan applied but post-apply validation failed
# Review the decision tree to understand what was done $ agents plan tree 01HXRCF1G2H3I4J5K6L7M8N9O0╭─ Decision Tree ────────────────────────────────────────────────────────────────────────────────╮ │ Plan: 01HXRCF1G2H3I4J5K6L7M8N9O0 │ │ ├─ [prompt_definition] "Optimize database connection handling" │ │ ├─ [invariant_enforced] "All DB config changes must preserve existing pool limits" │ │ ├─ [strategy_choice] "Refactor connection pool configuration" (confidence: 0.91) │ │ ├─ [tool_invocation] write_file: src/db/pool_config.py │ │ ├─ [implementation_choice] "Increase pool size to 100" (confidence: 0.78) ← ROOT CAUSE │ │ └─ [tool_invocation] write_file: src/db/health_check.py │ ╰────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Tree Summary ──────────────────╮ │ Nodes: 6 │ │ Depth: 2 │ │ Child Plans: 0 │ │ Invariants: 1 │ │ Superseded: 0 (hidden) │ ╰─────────────────────────────────╯
╭─ Decision IDs (for correction) ──────────────────────╮ │ Root: 01HXRCC1D2E3F4G5H6I7J8K9 │ │ Invariant 1: 01HXRCC2E3F4G5H6I7J8K9L0 │ │ Strategy: 01HXRCC3F4G5H6I7J8K9L0M1 │ │ Tool 1: 01HXRCC4G5H6I7J8K9L0M1N2 │ │ Pool Size (root cause): 01HXRCD3E4F5G6H7I8J9K0L1 │ │ Tool 2: 01HXRCE4F5G6H7I8J9K0L1M2 │ ╰──────────────────────────────────────────────────────╯
✓ OK Decision tree rendered
# Look at the specific changes that were applied $ agents plan diff 01HXRCF1G2H3I4J5K6L7M8N9O0╭─ Diff Summary ─────────────────────────────────────╮ │ Plan: 01HXRCF1G2H3I4J5K6L7M8N9O0 │ │ Project: local/api-service │ │ Files Changed: 4 │ │ Insertions: 38 │ │ Deletions: 12 │ │ Net Change: +26 lines │ ╰────────────────────────────────────────────────────╯
╭─ Files ──────────────────────────────────────╮ │ Path Change Status │ │ ──────────────────────── ────── ──────── │ │ src/db/pool_config.py +14 -6 modified │ │ src/db/health_check.py +18 -0 new file │ │ src/db/__init__.py +4 -4 modified │ │ config/database.toml +2 -2 modified │ ╰──────────────────────────────────────────────╯
╭─ Patch Preview ──────────────────────────────────╮ │ --- a/src/db/pool_config.py │ │ +++ b/src/db/pool_config.py │ │ @@ -8,6 +8,14 @@ │ │ - POOL_SIZE = 20 │ │ - POOL_MAX_OVERFLOW = 10 │ │ + POOL_SIZE = 100 │ │ + POOL_MAX_OVERFLOW = 50 │ │ + POOL_RECYCLE = 3600 │ │ + POOL_PRE_PING = True │ │ --- a/config/database.toml │ │ +++ b/config/database.toml │ │ @@ -3,2 +3,2 @@ │ │ - pool_size = 20 │ │ + pool_size = 100 │ ╰──────────────────────────────────────────────────╯
╭─ Risk Assessment ──────────────────────────╮ │ API Compatibility: preserved │ │ Test Coverage: maintained │ │ Breaking Changes: pool size 5x increase │ │ Resource Impact: high (connection limit) │ ╰────────────────────────────────────────────╯
✓ OK Diff generated
# Examine the decision that led to the problematic change $ agents plan explain 01HXRCD3E4F5G6H7I8J9K0L1 --show-context --show-reasoning╭─ Decision ───────────────────────────────────────────────────────────╮ │ ID: 01HXRCD3E4F5G6H7I8J9K0L1 │ │ Type: implementation_choice │ │ Question: What pool size should be configured? │ │ Chosen: Increase pool size from 20 to 100 │ │ Confidence: 0.78 │ │ Plan: 01HXRCF1G2H3I4J5K6L7M8N9O0 │ │ Sequence: 5 of 6 │ │ Created: 2026-02-11 08:38 │ ╰──────────────────────────────────────────────────────────────────────╯
╭─ Alternatives Considered ─────────────────────────────────────╮ │ 1. Increase pool size from 20 to 100 (chosen) │ │ 2. Keep pool at 20, add connection health checks only │ │ 3. Increase to 50 with overflow to 75 │ ╰───────────────────────────────────────────────────────────────╯
╭─ Impact ─────────────────────────╮ │ Downstream Decisions: 1 │ │ Downstream Child Plans: 0 │ │ Artifacts Produced: 2 │ │ Correction Impact: low │ ╰──────────────────────────────────╯
╭─ Context Snapshot ────────────────────────────────────────────────╮ │ - Current pool_size=20 in config/database.toml │ │ - Database server: PostgreSQL 15, max_connections=150 │ │ - 3 application replicas running (total: 60 connections used) │ │ - Slow query logs show connection wait times of 200-500ms │ │ Hot Context Hash: sha256:9c4f...a1b2 │ ╰───────────────────────────────────────────────────────────────────╯
╭─ Rationale ─────────────────────────────────────────────────────╮ │ Connection wait times of 200-500ms suggest pool exhaustion. │ │ Increasing to 100 connections per replica provides headroom │ │ for burst traffic and eliminates the queueing bottleneck. │ ╰─────────────────────────────────────────────────────────────────╯
╭─ Model Reasoning (raw) ────────────────────────────────────────────────╮ │ The slow query logs show connection acquisition delays of 200-500ms │ │ which indicates the pool is frequently exhausted. Current config: │ │ pool_size=20, max_overflow=10 (total 30 per replica). │ │ │ │ With 3 replicas that's 90 connections max. The PostgreSQL server │ │ allows 150. I'll increase to 100 per replica (300 total with │ │ overflow). This exceeds the server's max_connections=150 but │ │ overflow connections are temporary and unlikely to all be active │ │ simultaneously. │ │ │ │ [NOTE: The model failed to account for 3 replicas × 100 = 300 │ │ exceeding PostgreSQL max_connections=150] │ ╰────────────────────────────────────────────────────────────────────────╯
╭─ Correction ─────────────────────────────────────────────────╮ │ agents plan correct 01HXRCD3E4F5G6H7I8J9K0L1 │ │ --mode revert --guidance "Keep pool at 20, add health..." │ ╰──────────────────────────────────────────────────────────────╯
✓ OK Decision explained
Step 2: Roll Back to a Checkpoint
# List available checkpoints for this plan $ agents plan artifacts 01HXRCF1G2H3I4J5K6L7M8N9O0╭─ Artifacts ────────────────────────────────────────────────────────╮ │ Path Type Size Change Child Plan │ │ ─────────────────────── ───── ────── ─────── ────────── │ │ src/db/pool_config.py edit 1.4 KB +14 -6 root │ │ src/db/health_check.py write 2.1 KB +18 -0 root │ │ src/db/__init__.py edit 0.8 KB +4 -4 root │ │ config/database.toml edit 0.3 KB +2 -2 root │ ╰────────────────────────────────────────────────────────────────────╯
╭─ Summary ───────────╮ │ Total: 4 │ │ Writes: 1 (new) │ │ Edits: 3 (modified) │ │ Deletes: 0 │ │ Total Size: 4.6 KB │ ╰─────────────────────╯
╭─ Checkpoints ──────────────────────────────────────────────────────╮ │ ID Label Created │ │ ──────────────────── ─────────────────────────── ────────────── │ │ checkpoint_01HXRCP1 before pool config changes 08:36:12 │ │ checkpoint_01HXRCP2 before health check addition 08:39:45 │ │ checkpoint_01HXRCP3 pre-apply snapshot 08:42:00 │ ╰────────────────────────────────────────────────────────────────────╯
╭─ By Plan ──────────────╮ │ Root Plan: 4 artifacts │ ╰────────────────────────╯
✓ OK 4 artifacts listed
# Roll back to the checkpoint before the problematic pool config change $ agents plan rollback --yes 01HXRCF1G2H3I4J5K6L7M8N9O0 checkpoint_01HXRCP1╭─ Rollback Summary ─────────────────────────────────╮ │ Plan: 01HXRCF1G2H3I4J5K6L7M8N9O0 │ │ Checkpoint: checkpoint_01HXRCP1 │ │ Label: before pool config changes │ │ Files: 4 reverted │ ╰────────────────────────────────────────────────────╯
╭─ Changes Reverted ──────────────────────╮ │ File Action │ │ ──────────────────────── ────────── │ │ src/db/pool_config.py restored │ │ src/db/health_check.py removed │ │ src/db/__init__.py restored │ │ config/database.toml restored │ ╰─────────────────────────────────────────╯
╭─ Impact ──────────────────────────────────╮ │ Child Plans Invalidated: 0 │ │ Sandbox: restored to checkpoint_01HXRCP1 │ │ Decisions After CP: 2 discarded │ │ Tool Calls After CP: 3 undone │ ╰───────────────────────────────────────────╯
╭─ Post-Rollback State ──────────╮ │ Phase: execute │ │ State: queued (awaiting input) │ │ Checkpoints Remaining: 1 │ ╰────────────────────────────────╯
✓ OK Rollback complete
Step 3: Correct and Re-Execute
# Correct the decision that caused the issue $ agents plan correct 01HXRCD3E4F5G6H7I8J9K0L1 \ --mode revert \ --guidance "The connection pool size change caused the outage. \ Keep the pool at 20 connections maximum and add \ connection health checks instead." --yes╭─ Correction ──────────────────────────────────────────────╮ │ Mode: revert │ │ Impact: 1 decision, 0 child plans, 2 artifacts │ │ New Decision: 01HXRCCORR1A2B3C4D5 │ │ Corrects: 01HXRCD3E4F5G6H7I8J9K0L1 │ │ Attempt: 2 │ ╰───────────────────────────────────────────────────────────╯
╭─ Affected Subtree ──────────────╮ │ Decisions Invalidated: 1 │ │ Child Plans Rolled Back: 0 │ │ Artifacts Archived: 2 │ │ Unaffected Decisions: 4 │ ╰─────────────────────────────────╯
╭─ Sandbox Rollback ──────────────────╮ │ Checkpoint: checkpoint_01HXRCP1 │ │ Files Reverted: 2 │ │ Status: restored │ ╰─────────────────────────────────────╯
╭─ Recompute ───────────╮ │ Queued: 1 re-execute │ │ ETA: 2m │ ╰───────────────────────╯
╭─ History ───────────────────────────────────────────────╮ │ - Original decision superseded │ │ - Prior artifacts archived for comparison │ │ - agents plan diff --correction 01HXRCCORR1A2B3C4D5 │ ╰─────────────────────────────────────────────────────────╯
✓ OK Correction applied — re-executing affected subtree
# The affected subtree is re-executed. Review the new changes. $ agents plan diff --correction 01HXRCCORR1A2B3C4D5╭─ Correction Diff ────────────────────────────────────╮ │ Correction: 01HXRCCORR1A2B3C4D5 │ │ Original Decision: 01HXRCD3E4F5G6H7I8J9K0L1 │ │ Mode: revert │ │ Files Changed: 3 │ │ New Insertions: 32 │ │ New Deletions: 4 │ ╰──────────────────────────────────────────────────────╯
╭─ Comparison ─────────────────────────────────────────────────────────────╮ │ File Before (original) After (corrected) │ │ ──────────────────────── ────────────────── ────────────────────── │ │ src/db/pool_config.py +14 -6 (pool=100) +8 -2 (pool=20, checks) │ │ src/db/health_check.py +18 -0 (basic) +24 -0 (robust checks) │ │ config/database.toml +2 -2 (pool=100) (unchanged — pool=20) │ ╰──────────────────────────────────────────────────────────────────────────╯
╭─ Patch Preview (corrected vs original) ───────────────────────╮ │ --- a/src/db/pool_config.py (original) │ │ +++ b/src/db/pool_config.py (corrected) │ │ @@ -8,6 +8,10 @@ │ │ - POOL_SIZE = 100 │ │ - POOL_MAX_OVERFLOW = 50 │ │ + POOL_SIZE = 20 │ │ + POOL_MAX_OVERFLOW = 10 │ │ + POOL_PRE_PING = True │ │ + POOL_HEALTH_CHECK_INTERVAL = 30 │ │ --- a/src/db/health_check.py (original) │ │ +++ b/src/db/health_check.py (corrected) │ │ @@ -1,18 +1,24 @@ │ │ + # Added: connection validation, stale connection eviction, │ │ + # periodic health probes, and circuit breaker pattern │ │ ... │ ╰───────────────────────────────────────────────────────────────╯
✓ OK Correction diff generated
# Re-apply with the fix $ agents plan apply --yes 01HXRCF1G2H3I4J5K6L7M8N9O0╭─ Apply Summary ──────────────────────────╮ │ Plan: 01HXRCF1G2H3I4J5K6L7M8N9O0 │ │ Artifacts: 3 files updated │ │ Changes: 32 insertions, 4 deletions │ │ Project: local/api-service │ │ Applied At: 2026-02-11 09:02 │ ╰──────────────────────────────────────────╯
╭─ Validation ───────────────────╮ │ Tests: passed (31/31) │ │ Lint: passed (0 warnings) │ │ Type Check: passed (0 errors) │ │ Health Check: passed │ │ Duration: 18.6s │ ╰────────────────────────────────╯
╭─ Sandbox Cleanup ─────────╮ │ Worktree: removed │ │ Branch: merged to main │ │ Checkpoint: archived │ ╰───────────────────────────╯
╭─ Plan Lifecycle ──────────────────────────╮ │ Phase: apply │ │ State: applied │ │ Total Duration: 00:32:02 (incl. rollback) │ │ Total Cost: $0.118 │ │ Decisions Made: 7 (1 corrected) │ │ Child Plans: 0 │ ╰───────────────────────────────────────────╯
╭─ Next Steps ──────╮ │ - Review git diff │ │ - Commit changes │ ╰───────────────────╯
✓ OK Changes applied
Architecture
This section provides the complete architectural blueprint for CleverAgents. It defines the system's structural composition, technology choices, layer boundaries, data flow, and operational characteristics. An implementor should be able to build the entire system from this section combined with the behavioral specification above.
Architecture Overview
!!! adr "Architecture Decision" The layered architecture, DDD boundaries, and deployment modes are defined in ADR-001: Layered Architecture.
CleverAgents is architected as a layered, event-driven application following Domain-Driven Design (DDD) principles with clean architecture boundaries. The system is organized into four primary layers — Presentation, Application, Domain, and Infrastructure — with strict dependency rules: outer layers depend on inner layers, never the reverse. Cross-cutting concerns (logging, configuration, security) are handled through dependency injection and aspect-oriented patterns.
The architecture supports two deployment modes:
- Local Mode: A single-process CLI application where all components run in the same Python process. The CLI is the primary entry point, the database is SQLite, and all resources are local.
- Server Mode: A multi-user service where the CLI acts as a thin client communicating with a remote server over HTTPS. The server hosts shared storage, namespace resolution, and remote plan execution.
Both modes share the same domain model and application layer. The infrastructure layer provides swappable implementations for each deployment target.
High-Level Component Diagram
@startuml
skinparam componentStyle rectangle
skinparam defaultFontSize 12
skinparam packageStyle frame
skinparam packageFontSize 14
skinparam packageFontStyle bold
skinparam componentFontSize 11
package "Presentation Layer" <<Frame>> #E0F7FA {
component [CLI\n(Typer)] as CLI
component [TUI\n(Textual)] as TUI
component [Web\n(Textual Web)] as Web
component [IDE Plugin\n(Language Server)] as IDE
component [REST API\n(uvicorn / FastAPI)] as REST
}
package "Application Layer" <<Frame>> #E3F2FD {
package "Service Facade" {
component [PlanService] as PS
component [ProjectService] as ProjS
component [ActorService] as ActS
component [ContextService] as CtxS
component [ToolService] as ToolS
component [SkillService] as SkillS
component [ResourceSvc] as ResS
}
package "Workflow Engine" {
component [PlanLifecycle] as PL
component [SessionWorkflow] as SW
component [CorrectionFlow] as CF
component [MergeWorkflow] as MW
}
package "Event Bus" {
component [StructuredLog] as SL
component [EventEmitter] as EE
component [MetricsCollector] as MC
}
component [DI Container\n(dependency-injector DeclarativeContainer)] as DI #D1C4E9
}
package "Domain Layer" <<Frame>> #F3E5F5 {
package "Domain Models" {
component [Plan / Decision / Project\nResource / Actor / Action\nTool / Skill / Session\nInvariant / AutomationProf\nCheckpoint / CorrectionAtmpt] as DM
}
package "Domain Services" {
component [InvariantEnforcer\nAutonomyCtrl\nContextBuilder\nMergeResolver] as DS
note bottom of DS : Repository Interfaces\n(Protocol classes)
}
package "Domain Events" {
component [PlanCreated / PhaseChanged\nDecisionMade / ToolInvoked\nApplyCompleted / CorrectionReq] as DE
}
}
package "Infrastructure Layer" <<Frame>> #FFEBEE {
package "Database" {
component [SQLite / SQLAlchemy\nAlembic / UoW] as DB
}
package "Indexing" {
component [Tantivy / FAISS\nNeo4j / Qdrant / RDFLib] as IDX
}
package "Sandbox" {
component [GitWorktree / FsCopy\nTxnRollback / NoOp] as SBX
}
package "LLM / AI Runtime" {
component [LangChain / LangGraph\nProviderRegistry\nRxPY Bridge / MCP SDK] as LLM
}
package "External Integrations" {
component [MCP Servers\nAgent Skills Std\nREST/gRPC Clients] as EXT
}
package "File System / OS" {
component [Watchdog / File I/O\nGit CLI / Subprocess\nOS Process / Signals] as FS
}
}
CLI -[hidden]right-> TUI
TUI -[hidden]right-> Web
Web -[hidden]right-> IDE
IDE -[hidden]right-> REST
CLI -down-> PS
TUI -down-> PS
Web -down-> PS
IDE -down-> PS
REST -down-> PS
PS -down-> DI
DI -down-> DM
DS -down-> DB
DS -down-> IDX
@enduml
Architectural Principles
-
Hexagonal Architecture (Ports and Adapters): The domain layer defines abstract repository interfaces (ports). The infrastructure layer provides concrete implementations (adapters). This allows swapping SQLite for PostgreSQL, FAISS for Qdrant, or local execution for remote server execution without touching domain logic.
-
Command Query Responsibility Segregation (CQRS): Write operations (plan creation, decision recording, resource modification) flow through command handlers that enforce invariants and emit domain events. Read operations (plan status, decision tree visualization, context queries) use optimized read paths that may bypass the ORM for performance.
-
Event-Driven Architecture: All significant state changes emit domain events through a structured event bus. Events flow through RxPY reactive streams for real-time processing and through the structured logging pipeline for persistence. This enables decoupled observability, audit logging, and future webhook/notification systems.
-
Dependency Injection: All service dependencies are wired through a
DeclarativeContainer(from thedependency-injectorlibrary). This eliminates hidden coupling, enables testing with mock implementations, and supports runtime reconfiguration (e.g., switching providers). -
Convention over Configuration: Default behavior requires zero configuration. For entity registration commands, YAML configuration files are the complete definition (no CLI overrides). For runtime commands (e.g.,
plan use), CLI flags may override entity defaults. Environment variables can be interpolated into configuration files via${ENV_VAR}syntax. The resolution chain depends on context: entity registration uses config-file-only semantics, while runtime commands support per-invocation CLI overrides of entity defaults.
Data Flow: Plan Lifecycle
The following diagram traces data flow through the system during a complete plan lifecycle:
sequenceDiagram
participant U as User
participant C as CLI
participant A as Application
participant D as Domain
participant I as Infrastructure
U->>C: plan use
C->>A: PlanService.create_plan()
A->>D: Plan(phase=strategize)
D->>I: PlanRepository.save(plan)
A->>D: InvariantEnforcer.reconcile()
A->>D: AutonomyCtrl.should_proceed
rect rgb(200, 230, 255)
Note over A,D: Strategize Phase (if auto)
A->>D: StrategyActor.invoke()
D->>I: LangGraph StateGraph.invoke
D->>I: ContextBuilder.build_hot_ctx()
I-->>D: [FAISS/Tantivy results]
D->>I: LLM Provider.generate()
I-->>D: [LLM response]
D-->>A: Decision[]
D->>I: DecisionRepo.save_batch()
end
rect rgb(255, 230, 200)
Note over A,D: Execute Phase
A->>D: SandboxManager.create()
D->>I: GitWorktree.create_worktree()
A->>D: ExecutionActor.invoke(sandbox)
D->>I: Tool calls via SkillRegistry
D-->>A: ChangeSet
end
rect rgb(200, 255, 200)
Note over A,D: Apply Phase
A->>D: Validation.run_all()
D->>I: Subprocess.run(commands)
A->>D: SandboxManager.commit()
D->>I: git merge
end
C-->>U: OK Applied
Technical Stack
!!! adr "Architecture Decision" The technology selections, version constraints, and rationale are defined in ADR-005: Technical Stack.
This section enumerates every technology choice in the CleverAgents stack, organized by functional area. Version constraints are minimum versions; newer compatible versions are acceptable.
Core Runtime
| Technology | Version | Role | Rationale |
|---|---|---|---|
| Python | >= 3.13 | Primary language | Type hint maturity, async/await, pattern matching, performance improvements in 3.13. The entire codebase is Python-only for consistency and LLM ecosystem alignment. |
| Hatchling | >= 1.21.0 | PEP 517 build backend | Lightweight, standards-compliant build system. Produces wheels and sdists without complex configuration. |
| uv | >= 0.8.0 | Package installer and resolver | 10-100x faster than pip. Used in CI, Docker builds, and Nox sessions for dependency installation. Drop-in replacement for pip with full PEP 723 support. |
CLI and Presentation
| Technology | Version | Role | Rationale |
|---|---|---|---|
| Typer | >= 0.9.0 | CLI framework | Built on Click with automatic help generation, type inference from Python type hints, and Rich integration for styled output. Supports nested command groups matching the agents <noun> <verb> pattern. |
| Rich | (transitive via Typer) | Terminal rendering | Provides Panel, Table, Tree, Syntax, Progress, Spinner, and Markdown rendering used by all output formats except plain and structured (json/yaml). |
| Textual | (future) | TUI framework | Built by the Rich maintainers. Provides a reactive terminal UI framework that can be served as a web app via Textual Web, enabling the "single UI codebase" strategy for TUI, Web, and IDE plugin. |
LLM and AI Runtime
| Technology | Version | Role | Rationale |
|---|---|---|---|
| LangChain | >= 0.2.14 | LLM abstraction layer | Provider-agnostic interface for chat models, embeddings, output parsing, prompt templates, and tool calling. The BaseLanguageModel protocol enables swapping providers without changing application code. |
| LangGraph | (transitive) | Stateful workflow orchestration | StateGraph with conditional edges, checkpointing (MemorySaver), and streaming execution. Used for the plan generation graph (load_context -> analyze -> generate -> validate) and auto-debug graph. |
| LangChain Provider Packages | varies | LLM provider integrations | langchain-openai (>= 0.2.0), langchain-google-genai (>= 0.2.0), langchain-anthropic, langchain-groq, langchain-together, langchain-cohere. Each provides a ChatModel implementation. |
| LangChain Community | >= 0.2.14 | Community integrations | FAISS vector store, FakeListLLM/FakeEmbeddings for testing. |
| RxPY | >= 3.2.0 | Reactive stream processing | Subject, BehaviorSubject, ReplaySubject, and operators (map, filter, flat_map, debounce, throttle, scan) for real-time event routing between actors, stream-to-graph bridging, and backpressure management. |
| MCP SDK | >= 1.4.0 | Model Context Protocol | Client SDK for communicating with MCP servers. Enables CleverAgents to discover and invoke tools exposed by any MCP-compliant server. |
Data and Persistence
| Technology | Version | Role | Rationale |
|---|---|---|---|
| SQLite | (system) | Primary database | Zero-configuration, file-based, ACID-compliant. Sufficient for local mode; supports WAL mode for concurrent reads during plan execution. For server mode, the infrastructure layer swaps in PostgreSQL via the same SQLAlchemy interface. |
| SQLAlchemy | (transitive via Alembic) | ORM and database abstraction | Declarative model mapping, session management, Unit of Work pattern, and dialect abstraction enabling database portability. |
| Alembic | >= 1.13.1 | Database migrations | Version-controlled schema migrations with upgrade/downgrade support. Migrations are auto-applied on first run and during agents init. |
| python-ulid | >= 2.7.0 | ID generation | ULID generation for plans, decisions, resources, and correction attempts. ULIDs are lexicographically sortable by creation time, enabling efficient time-range queries without separate timestamp indexes. |
Indexing and Code Intelligence
| Technology | Version | Role | Rationale |
|---|---|---|---|
| Tantivy (via tantivy-py) | configurable | Full-text search | Rust-based search engine providing sub-millisecond full-text search across project resources. Used for keyword-based code search during context building. Alternative: SQLite FTS5 for zero-dependency installations. |
| FAISS (faiss-cpu) | >= 1.7.4 | Vector similarity search | Facebook AI Similarity Search for semantic code search. Indexes embedding vectors generated from project files. Supports approximate nearest neighbor search for fast retrieval during context building. |
| Qdrant | configurable | Vector search (alternative) | External vector database for production deployments requiring horizontal scaling, persistence, and filtering. Connected via qdrant-client. |
| Neo4j | configurable | Knowledge graph | Graph database for structural code relationships (class hierarchies, call graphs, dependency trees). Enables queries like "find all callers of function X" during Strategize. Alternative: rdflib for in-process graph queries. |
| rdflib | >= 7.1.4 | In-process graph store | Python RDF library for lightweight structural code analysis without external dependencies. Stores RDF triples representing code relationships. |
| OpenAI Embeddings | configurable | Vector generation | Default embedding provider (text-embedding-3-small). Alternative providers: Anthropic, local sentence-transformers models via index.embedding.provider config. |
Configuration and Validation
| Technology | Version | Role | Rationale |
|---|---|---|---|
| Pydantic | >= 2.7.0 | Data validation and modeling | All domain models, configuration objects, and API schemas use Pydantic v2 for runtime validation, JSON Schema generation, and serialization. Provides type-safe data boundaries between layers. |
| Pydantic Settings | >= 2.11.0 | Environment configuration | BaseSettings subclasses for environment variable loading with the CLEVERAGENTS_ prefix, .env file support, and nested model resolution. Implements the four-tier configuration resolution chain. |
| Jinja2 | >= 3.1.0 | Template rendering | Used for actor system prompts, YAML configuration interpolation, and prompt template rendering. Supports {{ context.* }} variable expansion with sandboxed execution. |
| TOML | (stdlib tomllib) | Configuration file format | Global configuration uses TOML for its native support of nested tables, which maps naturally to the dot-separated hierarchical key structure. Python 3.13 includes tomllib in stdlib. |
| YAML | (via PyYAML) | Entity configuration format | All entity definitions (actors, skills, tools, actions, resource types, automation profiles) use YAML as the canonical human-authored format. JSON is accepted as valid YAML. |
Testing
| Technology | Version | Role | Rationale |
|---|---|---|---|
| Behave | >= 1.2.6 | BDD feature tests | Gherkin .feature files describe behavioral scenarios in natural language. Step implementations exercise the full application stack. Custom behave-parallel runner enables parallel execution via ProcessPoolExecutor. |
| Robot Framework | >= 7.3.2 | Integration tests | Keyword-driven integration tests for end-to-end CLI workflows. robotframework-pabot enables parallel suite execution. |
| pytest | >= 8.0.0 | Unit tests | Standard Python test framework for unit-level testing with fixtures, parametrization, and plugin support. |
| pytest-asyncio | >= 0.23.0 | Async test support | Enables async def test_* functions for testing async LangGraph workflows and RxPY stream operations. |
| pytest-cov / coverage | >= 7.11.0 | Coverage reporting | Target: 85% line coverage. HTML and XML reports generated by Nox sessions. Coverage gates enforce minimum thresholds in CI. |
| Hypothesis | >= 6.136.6 | Property-based testing | Generates randomized test inputs for invariant verification, particularly useful for testing decision tree operations, merge strategies, and context tier boundaries. |
| ASV (Airspeed Velocity) | >= 0.6.5 | Performance benchmarks | Tracks plan generation latency, CLI startup time, and indexing throughput across commits. Prevents performance regressions. |
Code Quality
| Technology | Version | Role | Rationale |
|---|---|---|---|
| Ruff | >= 0.1.0 | Linting and formatting | Single tool replacing Black, Flake8, isort, and pyupgrade. Rules: E, F, W, B, UP, I, SIM, RUF. Line length: 88. Double quotes. 4-space indent. |
| Pyright | >= 1.1.350 | Static type checking | Strict mode type checking for the entire codebase. Catches type errors at development time rather than runtime. Configured via pyrightconfig.json. |
Infrastructure and Deployment
| Technology | Version | Role | Rationale |
|---|---|---|---|
| Docker | multi-stage | Containerization | Production image uses python:3.13-slim with multi-stage build. Non-root user (appuser, uid 1000). Entrypoint: python -m cleveragents. |
| Helm | (chart) | Kubernetes deployment | Helm chart in k8s/ directory for server mode deployment. Linted and template-tested in CI. |
| Nox | >= 2025.4.22 | Task automation | Session-based task runner for lint, test, build, docs, benchmarks, and coverage. Uses uv as the virtual environment backend for fast session creation. |
| Forgejo CI | (workflow) | Continuous integration | Self-hosted Forgejo instance with GitHub Actions-compatible workflow syntax. Pipeline: lint -> typecheck -> behave (matrix 3.11/3.12/3.13) -> build -> docker -> helm. |
Monitoring and Observability
| Technology | Version | Role | Rationale |
|---|---|---|---|
| structlog | >= 24.4.0 | Structured logging | JSON-structured log output with context binding (plan_id, decision_id, actor_name, tool_name). Enables log aggregation, filtering, and correlation across complex plan hierarchies. |
| LangSmith | (optional) | LLM observability | Optional integration for tracing LLM calls, token usage, latency, and cost. Configured via CLEVERAGENTS_LANGSMITH_* environment variables. Syncs to LANGCHAIN_TRACING_V2. |
Additional Libraries
| Technology | Version | Role | Rationale |
|---|---|---|---|
| dependency-injector | >= 4.41.0 | DI container | DeclarativeContainer with Singleton, Factory, and Configuration providers. Wires all services, repositories, and infrastructure components. |
| watchdog | >= 4.0.0 | File system monitoring | Watches project resource directories for changes. Triggers automatic re-indexing when index.auto-reindex is enabled. |
| numpy | >= 2.1.0 | Numerical computing | Vector operations for embedding similarity computation, confidence score aggregation, and distance calculations in the vector store layer. |
| uvicorn | >= 0.30.1 | ASGI server | HTTP server for server mode REST API. Serves the FastAPI application for remote plan execution, namespace resolution, and multi-user collaboration. |
Dependency Injection
!!! adr "Architecture Decision" The DI container design, wiring pattern, and testing strategy are defined in ADR-003: Dependency Injection.
All service dependencies in CleverAgents are wired through a DeclarativeContainer from the dependency-injector library (>= 4.41.0). This container is the single composition root for the entire application. No service instantiates its own dependencies — all are received through constructor injection managed by the container.
Container Structure
The DI container is a DeclarativeContainer subclass that declares all providers:
Singletonproviders for long-lived services: database engine, repository implementations, event bus, index backends, provider registry.Factoryproviders for per-request or per-operation objects: sandbox instances, plan execution contexts, LangGraph state graphs.Configurationprovider for loading and exposing the global TOML configuration and environment variables.
Wiring Pattern
The wiring pattern enforces the layered architecture (see Architecture Overview) by ensuring that dependencies always flow inward:
- The Domain Layer defines repository interfaces as
Protocolclasses (e.g.,PlanRepository,DecisionRepository,ResourceRepository). - The Infrastructure Layer provides concrete implementations (e.g.,
SQLAlchemyPlanRepository,SQLAlchemyDecisionRepository). - The DI container maps protocols to implementations using
SingletonorFactoryproviders. - Application Layer services declare their dependencies as constructor parameters typed to the domain protocols.
- The container auto-wires these dependencies at application startup.
This pattern ensures that domain and application code never imports infrastructure modules directly. Swapping backends (e.g., SQLite → PostgreSQL, FAISS → Qdrant, local execution → remote server) requires only changing the container provider mapping — no application or domain code changes.
Provider Types
| Provider Type | Use Case | Example |
|---|---|---|
Singleton |
Shared stateful services | Database session factory, event bus, index backends |
Factory |
Per-use transient objects | Sandbox instances, execution contexts |
Configuration |
External configuration values | TOML config keys, environment variables |
Testing Support
In test configurations, the container is reconfigured to supply mock or fake implementations:
FakeListLLM/FakeEmbeddingsfrom LangChain Community replace real LLM providers.- In-memory repositories replace SQLAlchemy-backed repositories.
- No-op sandbox strategies replace git worktree or filesystem copy strategies.
The container's override() context manager allows swapping providers for individual test cases without affecting the global configuration.
Container Rules
- All cross-layer dependencies must flow through the DI container. Direct instantiation of infrastructure components in application or domain code is prohibited.
- The container is initialized exactly once at application startup (in the CLI entry point or server bootstrap). It must not be re-initialized during request handling.
- Domain Layer code must never import or reference the
dependency-injectorlibrary. Domain protocols are plain PythonProtocolclasses with no DI framework annotations. - Every new service or repository must be registered in the container before use. Unregistered dependencies cause explicit startup failures, not silent runtime errors.
- A startup self-test verifies that all declared providers can be resolved without errors. This runs as part of the application bootstrap and in CI.
Data Validation
!!! adr "Architecture Decision" The data validation strategy, Pydantic usage, and schema generation approach are defined in ADR-004: Data Validation.
CleverAgents processes data from multiple untrusted or semi-trusted sources: user CLI input, YAML configuration files, LLM-generated tool call arguments, MCP server responses, environment variables, and database records. Every boundary crossing — between layers, between the system and external services, between the user and the application — is an opportunity for invalid data to enter the system and cause subtle downstream failures.
All domain models, configuration objects, and API schemas use Pydantic V2 (>= 2.7.0) for runtime data validation, JSON Schema generation, and serialization. Pydantic Settings (>= 2.11.0) extends this to environment variable loading. Pydantic serves as the type-safe data boundary between all layers.
Domain Models
Every domain entity — Plan, Decision, Action, Resource, Actor, Tool, Skill, Session, Invariant, AutomationProfile, Checkpoint, CorrectionAttempt — is a Pydantic BaseModel subclass. Fields use Python type annotations with Pydantic field validators for business rule enforcement.
Configuration Objects
YAML entity configuration files (actors, skills, tools, actions, resource types, automation profiles, context views) are parsed into Pydantic models. Validation errors are reported with field paths and human-readable messages before any processing occurs. A partially valid configuration must never be applied — YAML files must validate completely before any side effects occur.
Environment and Global Configuration
BaseSettings subclasses from Pydantic Settings handle environment variable loading with the CLEVERAGENTS_ prefix, .env file support, and nested model resolution. This implements the four-tier configuration resolution chain: CLI flag > environment variable > project-scoped config > global config file > built-in default.
JSON Schema Generation
Pydantic models automatically generate JSON Schema definitions, which are used for:
- Tool input/output schema definitions consumed by LLM tool-calling protocols.
- Validation of LLM-generated tool call arguments before execution.
- API request/response schema documentation.
JSON Schema generated from Pydantic models is the canonical schema for tool definitions. Manual schema definitions must not diverge from the model.
Serialization
Pydantic handles serialization to and from JSON, YAML (via dict intermediary), and database column types. Custom serializers are defined for domain-specific types (ULIDs, datetime formats, enum values).
Data Validation Rules
- All data crossing a layer boundary must pass through a Pydantic model. Raw dictionaries or untyped data structures must not propagate between layers.
- Strict mode is used where appropriate to prevent implicit type coercion (e.g., strings silently becoming integers).
- Validation error messages must include the field path and a human-readable explanation suitable for CLI display.
- Pyright in strict mode verifies that Pydantic model field types are consistent across the codebase.
- Schema drift tests verify that JSON Schema generated from Pydantic tool models matches the schemas expected by MCP and LangChain tool-calling protocols.
ACMS (Advanced Context Management System)
!!! adr "Architecture Decision" The ACMS architecture, UKO layer, CRP protocol, and context strategy framework are defined in ADR-014: Context Management (ACMS).
This section provides the complete architectural design of the Advanced Context Management System (ACMS). For the conceptual overview, core data types, CRP skill definition, strategy protocol, and plan lifecycle integration, see Core Concepts > Advanced Context Management System (ACMS).
Architectural Overview
@startuml
skinparam componentStyle rectangle
skinparam defaultFontSize 12
skinparam packageFontSize 13
skinparam packageFontStyle bold
skinparam componentFontSize 11
package "Actor / Skill" as actor #E0F7FA {
component [Issues ContextRequests\nvia Context Request Protocol] as CRP
}
package "Context Assembly Pipeline\n(10-component pluggable)" as pipeline #C8E6C9 {
package "Phase 1: Strategy Orchestration" {
component [StrategySelector] as SS
component [BudgetAllocator] as BA
component [StrategyExecutor] as SE
}
package "Phase 2: Fragment Fusion" {
component [Deduplicator → DepthResolver\n→ Scorer → Packer → Orderer] as Fusion
}
package "Phase 3: Finalization" {
component [PreambleGenerator\nSkeletonCompressor] as Final
}
}
package "Strategies" as strats #E8EAF6 {
component [ARCE\n(graph-aware)] as S1
component [Simple\n(keyword/embed)] as S2
component [Breadth/Depth\nNavigator] as S3
component [Custom\n(user-defined)] as S4
}
package "Backend Abstraction Layer (BAL)" as bal #FFF9C4 {
component [ScopedView\n(Text DB)] as TV
component [ScopedView\n(Vector DB)] as VV
component [ScopedView\n(Graph DB)] as GV
}
package "Physical Data Stores" as stores #E3F2FD {
component [Tantivy /\nSQLite FTS] as TDB
component [FAISS /\nQdrant / Weaviate] as VDB
component [Blazegraph / Jena /\nNeo4j / Stardog\n(UKO triples)] as GDB
}
package "Resource Registry" as rr #F3E5F5 {
component [Physical/virtual resources\nDAG links] as RR
}
CRP -down-> SS
SS -down-> BA
BA -down-> SE
SE -down-> S1
SE -down-> S2
SE -down-> S3
SE -down-> S4
S1 -down-> TV
S1 -down-> VV
S1 -down-> GV
S2 -down-> TV
S2 -down-> VV
S3 -down-> GV
S4 -down-> TV
SE -right[hidden]-> Fusion
Fusion -right[hidden]-> Final
TV -down-> TDB
VV -down-> VDB
GV -down-> GDB
TDB -down-> RR
VDB -down-> RR
GDB -down-> RR
@enduml
Key Design Principles:
- Strategy-agnostic framework: The framework itself has no opinion about how context is assembled. That is the job of strategies. The framework provides the request protocol, the data plumbing, and the Context Assembly Pipeline.
- Dynamic budget: The token budget for context assembly may change on every invocation. Strategies receive the current budget and must respect it. The pipeline's BudgetPacker enforces the budget as a hard ceiling.
- Hierarchical by default: Every context assembly occurs within the context of a plan hierarchy. The system tracks parent context and provides progressive focusing primitives.
- Provenance everywhere: Every piece of context delivered to an actor can be traced back to a specific resource, a specific location within that resource, a specific revision, and a specific UKO node.
- Pluggable Architecture: Every component can be extended or replaced.
- Progressive Enhancement: System works with basic text search, enhances with advanced features.
- Eager Indexing: Indices are built immediately when resources are added and kept continuously up-to-date.
- Agent Awareness: Agents understand available indices through skills.
- Real-time Synchronization: Indices update immediately as code changes.
Universal Knowledge Ontology (UKO)
Design Philosophy
The UKO is designed around a single principle: everything that an actor might want to know about a resource should be representable as a node in the knowledge graph, at any level of detail, with provenance back to the originating resource.
The ontology uses an inheritance hierarchy so that:
- Layer 0 defines universal concepts that apply to any information (code, docs, data, infra). DetailDepth is defined here as a non-negative integer with domain-agnostic semantics: depth 0 is the most minimal representation (just a name/identifier), and each increment reveals progressively more structure, content, and relationships. There is no fixed upper bound — the maximum meaningful depth depends on the domain.
- Layer 1 specializes for domain-specific concepts: general software (
uko-code:— modules, callables, types), documents (uko-doc:— sections, paragraphs, citations), data schemas (uko-data:— tables, columns, constraints), and infrastructure (uko-infra:— services, endpoints, config). Each domain registers a DetailLevelMap that assigns named labels to specific integer depths (e.g.,uko-code:mapsMODULE_LISTING→ 0,SIGNATURES→ 4,FULL_SOURCE→ 9). - Layer 2 specializes for paradigm/format-specific concepts within a domain: object-oriented (
uko-oo:), functional (uko-func:), procedural (uko-proc:) for code. Future extensions could add academic vs. technical for documents, or relational vs. graph for databases. Each paradigm extends its parent domain's DetailLevelMap, potentially inserting new named levels between existing ones (which shifts subsequent integer assignments). - Layer 3 specializes for technology-specific concepts: Python (
uko-py:), TypeScript (uko-ts:), Rust (uko-rs:), Java (uko-java:), and potentially Markdown, PostgreSQL, or others. Language/technology-specific DetailLevelMap extensions are added only when they provide meaningful additional semantics beyond the paradigm level.
This layering means a strategy that operates at Layer 0 can work with any resource — code, documents, databases, infrastructure — while a strategy that operates at Layer 3 can provide language-specific or technology-specific intelligence. The breadth-depth-navigator strategy typically operates at Layers 0-1 (universal graph traversal), while arce leverages all layers. DetailDepth resolution follows the DetailLevelMap inheritance chain: when a named level is requested, the system looks up the name in the most specific map available for the target UKO node type (e.g., uko-py: first), then walks up to the parent map (uko-oo: → uko-code: → uko:) until a match is found. When a raw integer is specified, it is used directly.
Ontology Hierarchy
The UKO defines five namespace prefixes:
| Prefix | IRI | Purpose |
|---|---|---|
uko: |
https://cleveragents.ai/ontology/uko# |
Layer 0: Universal foundation |
uko-code: |
https://cleveragents.ai/ontology/uko/code# |
Layer 1: General software |
uko-oo: |
https://cleveragents.ai/ontology/uko/oo# |
Layer 2: Object-oriented paradigm |
uko-func: |
https://cleveragents.ai/ontology/uko/func# |
Layer 2: Functional paradigm |
uko-proc: |
https://cleveragents.ai/ontology/uko/proc# |
Layer 2: Procedural paradigm |
uko-py: |
https://cleveragents.ai/ontology/uko/py# |
Layer 3: Python-specific |
uko-ts: |
https://cleveragents.ai/ontology/uko/ts# |
Layer 3: TypeScript-specific |
uko-rs: |
https://cleveragents.ai/ontology/uko/rs# |
Layer 3: Rust-specific |
uko-java: |
https://cleveragents.ai/ontology/uko/java# |
Layer 3: Java-specific |
uko-doc: |
https://cleveragents.ai/ontology/uko/doc# |
Layer 1: Documents |
uko-data: |
https://cleveragents.ai/ontology/uko/data# |
Layer 1: Data schemas |
uko-infra: |
https://cleveragents.ai/ontology/uko/infra# |
Layer 1: Infrastructure |
Layer 0: Universal Foundation (uko:)
Every UKO node, regardless of domain, is one of five base classes:
@prefix uko: <https://cleveragents.ai/ontology/uko#> . @prefix owl: <http://www.w3.org/2002/07/owl#> . @prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> . @prefix xsd: <http://www.w3.org/2001/XMLSchema#> .# ── Base Classes ────────────────────────────────────────────────── uko:InformationUnit a owl:Class ; rdfs:comment "The root of every UKO node. Anything that can appear in an actor's context." .
uko:Container a owl:Class ; rdfs:subClassOf uko:InformationUnit ; rdfs:comment "An information unit that contains other information units (file, module, class, section)." .
uko:Atom a owl:Class ; rdfs:subClassOf uko:InformationUnit ; rdfs:comment "A leaf-level information unit (function body, paragraph, config value)." .
uko:Annotation a owl:Class ; rdfs:subClassOf uko:InformationUnit ; rdfs:comment "Metadata attached to another information unit (comment, docstring, attribute)." .
uko:Boundary a owl:Class ; rdfs:subClassOf uko:InformationUnit ; rdfs:comment "An interface point between containers (export, API endpoint, public method signature)." .
# ── Core Relationships ──────────────────────────────────────────── uko:contains a owl:ObjectProperty ; rdfs:domain uko:Container ; rdfs:range uko:InformationUnit ; rdfs:comment "Parent contains child (e.g., module contains class)." .
uko:references a owl:ObjectProperty ; rdfs:domain uko:InformationUnit ; rdfs:range uko:InformationUnit ; rdfs:comment "Weak reference (e.g., a function mentions a type in a docstring)." .
uko:dependsOn a owl:ObjectProperty ; rdfs:domain uko:InformationUnit ; rdfs:range uko:InformationUnit ; rdfs:comment "Strong dependency (e.g., import, inheritance, call)." .
# ── Content Properties ──────────────────────────────────────────── uko:hasRendering a owl:DatatypeProperty ; rdfs:domain uko:InformationUnit ; rdfs:range xsd:string ; rdfs:comment "A rendered text representation of this node at a specific detail depth. Multiple renderings at different depths may exist for the same node. The depth is indicated by the associated uko:renderingDepth value." .
uko:renderingDepth a owl:DatatypeProperty ; rdfs:domain uko:InformationUnit ; rdfs:range xsd:nonNegativeInteger ; rdfs:comment "The integer detail depth at which the associated uko:hasRendering was produced. Paired with uko:hasRendering via a blank node or reification." .
uko:hasFullContent a owl:DatatypeProperty ; rdfs:domain uko:InformationUnit ; rdfs:range xsd:string ; rdfs:comment "Shorthand for the maximum-depth rendering of this node (complete, unabridged content). Equivalent to uko:hasRendering at the domain's maximum depth." .
# ── Provenance ──────────────────────────────────────────────────── uko:sourceResource a owl:ObjectProperty ; rdfs:domain uko:InformationUnit ; rdfs:comment "The CleverAgents Resource (by ULID) this node was extracted from." .
uko:sourcePath a owl:DatatypeProperty ; rdfs:domain uko:InformationUnit ; rdfs:range xsd:string ; rdfs:comment "Path within the resource (e.g., file path relative to the resource root)." .
uko:sourceRange a owl:DatatypeProperty ; rdfs:domain uko:InformationUnit ; rdfs:range xsd:string ; rdfs:comment "Byte or line range within the source file (e.g., '42:1-87:0')." .
# ── Temporal ────────────────────────────────────────────────────── uko:validFrom a owl:DatatypeProperty ; rdfs:domain uko:InformationUnit ; rdfs:range xsd:dateTime .
uko:validUntil a owl:DatatypeProperty ; rdfs:domain uko:InformationUnit ; rdfs:range xsd:dateTime .
uko:isCurrent a owl:DatatypeProperty ; rdfs:domain uko:InformationUnit ; rdfs:range xsd:boolean .
uko:isRevisionOf a owl:ObjectProperty ; rdfs:domain uko:InformationUnit ; rdfs:range uko:InformationUnit ; rdfs:comment "Links a revised node to its predecessor." .
Layer 1: Domain Specializations
Layer 1 specializes the universal Layer 0 foundation for four domains: general software (uko-code:), documents (uko-doc:), data schemas (uko-data:), and infrastructure (uko-infra:). Each domain defines its own classes, relationships, properties, and a DetailLevelMap mapping named levels to integer DetailDepth values.
General Software (uko-code:):
@prefix uko-code: <https://cleveragents.ai/ontology/uko/code#> .uko-code:Module a owl:Class ; rdfs:subClassOf uko:Container ; rdfs:comment "A source code module (file, package, namespace)." .
uko-code:Callable a owl:Class ; rdfs:subClassOf uko:Atom ; rdfs:comment "Any callable unit (function, method, procedure, lambda)." .
uko-code:TypeDefinition a owl:Class ; rdfs:subClassOf uko:Atom ; rdfs:comment "A type or schema definition (class, struct, interface, enum, typedef)." .
uko-code:TestCase a owl:Class ; rdfs:subClassOf uko:Atom ; rdfs:comment "A test case or test function." .
uko-code:Import a owl:Class ; rdfs:subClassOf uko:Annotation ; rdfs:comment "An import/include/require statement." .
uko-code:hasReturnType a owl:DatatypeProperty ; rdfs:domain uko-code:Callable ; rdfs:range xsd:string .
uko-code:hasParameters a owl:DatatypeProperty ; rdfs:domain uko-code:Callable ; rdfs:range xsd:string ; rdfs:comment "JSON-encoded parameter list." .
uko-code:testsCallable a owl:ObjectProperty ; rdfs:domain uko-code:TestCase ; rdfs:range uko-code:Callable ; rdfs:comment "Links a test case to the callable it tests." .
Documents (uko-doc:) — Layer 1:
The document ontology represents structured and semi-structured documents (articles, technical papers, documentation, reports) as a hierarchy of containers and atoms. A key design principle is semantic awareness: when the system processes a document, it does not merely parse the structural hierarchy (sections, subsections, paragraphs) — it also analyzes the content of each text unit to infer implicit relationships. If a paragraph discusses a concept that is the subject of another section (even without an explicit hyperlink or mention by name), the RDF graph captures that relationship via uko:references or uko-doc:discussesTopic edges. This enables strategies to surface related content that a purely structural traversal would miss.
@prefix uko-doc: <https://cleveragents.ai/ontology/uko/doc#> .# ── Container Classes ────────────────────────────────────────────── uko-doc:Document a owl:Class ; rdfs:subClassOf uko:Container ; rdfs:comment "A complete document (article, report, manual, specification)." .
uko-doc:Part a owl:Class ; rdfs:subClassOf uko:Container ; rdfs:comment "A top-level division of a document (e.g., Part I, Part II)." .
uko-doc:Chapter a owl:Class ; rdfs:subClassOf uko:Container ; rdfs:comment "A chapter within a document or part." .
uko-doc:Section a owl:Class ; rdfs:subClassOf uko:Container ; rdfs:comment "A numbered or titled section within a chapter." .
uko-doc:Subsection a owl:Class ; rdfs:subClassOf uko-doc:Section ; rdfs:comment "A subsection nested within a section (arbitrary depth)." .
# ── Atom Classes ─────────────────────────────────────────────────── uko-doc:Paragraph a owl:Class ; rdfs:subClassOf uko:Atom ; rdfs:comment "A paragraph of prose text." .
uko-doc:CodeBlock a owl:Class ; rdfs:subClassOf uko:Atom ; rdfs:comment "An inline code listing or example within a document." .
uko-doc:Figure a owl:Class ; rdfs:subClassOf uko:Atom ; rdfs:comment "A figure, diagram, or image with an optional caption." .
uko-doc:Table a owl:Class ; rdfs:subClassOf uko:Atom ; rdfs:comment "A tabular data element within a document." .
uko-doc:BlockQuote a owl:Class ; rdfs:subClassOf uko:Atom ; rdfs:comment "A quoted passage from another source." .
uko-doc:ListBlock a owl:Class ; rdfs:subClassOf uko:Atom ; rdfs:comment "An ordered or unordered list." .
# ── Annotation Classes ───────────────────────────────────────────── uko-doc:Footnote a owl:Class ; rdfs:subClassOf uko:Annotation ; rdfs:comment "A footnote or endnote." .
uko-doc:Citation a owl:Class ; rdfs:subClassOf uko:Annotation ; rdfs:comment "A bibliographic citation or reference." .
uko-doc:Comment a owl:Class ; rdfs:subClassOf uko:Annotation ; rdfs:comment "An editorial comment or annotation (e.g., HTML comment, review note)." .
# ── Boundary Classes ─────────────────────────────────────────────── uko-doc:TableOfContents a owl:Class ; rdfs:subClassOf uko:Boundary ; rdfs:comment "The document's table of contents — an interface into the document's structure." .
uko-doc:Index a owl:Class ; rdfs:subClassOf uko:Boundary ; rdfs:comment "A back-of-book index or keyword index." .
uko-doc:Glossary a owl:Class ; rdfs:subClassOf uko:Boundary ; rdfs:comment "A glossary of terms defined in the document." .
# ── Document-Specific Relationships ──────────────────────────────── uko-doc:discussesTopic a owl:ObjectProperty ; rdfs:subPropertyOf uko:references ; rdfs:domain uko:InformationUnit ; rdfs:range uko:InformationUnit ; rdfs:comment "Semantic relationship: this unit discusses a topic that is the subject of another unit. Inferred by the document analyzer even when no explicit link exists." .
uko-doc:cites a owl:ObjectProperty ; rdfs:subPropertyOf uko:references ; rdfs:domain uko:InformationUnit ; rdfs:range uko-doc:Citation ; rdfs:comment "This unit cites a bibliographic reference." .
uko-doc:crossReferences a owl:ObjectProperty ; rdfs:subPropertyOf uko:references ; rdfs:domain uko:InformationUnit ; rdfs:range uko:InformationUnit ; rdfs:comment "Explicit cross-reference (e.g., 'see Section 3.2')." .
uko-doc:precedes a owl:ObjectProperty ; rdfs:domain uko:InformationUnit ; rdfs:range uko:InformationUnit ; rdfs:comment "Reading order: this unit comes before the target unit." .
# ── Document-Specific Properties ─────────────────────────────────── uko-doc:headingLevel a owl:DatatypeProperty ; rdfs:domain uko-doc:Section ; rdfs:range xsd:integer ; rdfs:comment "The nesting depth of this section (1 = top-level, 2 = subsection, etc.)." .
uko-doc:headingText a owl:DatatypeProperty ; rdfs:domain uko-doc:Section ; rdfs:range xsd:string ; rdfs:comment "The title/heading text of this section." .
uko-doc:wordCount a owl:DatatypeProperty ; rdfs:domain uko:InformationUnit ; rdfs:range xsd:integer ; rdfs:comment "Word count of the text content in this unit." .
uko-doc:language a owl:DatatypeProperty ; rdfs:domain uko-doc:Document ; rdfs:range xsd:string ; rdfs:comment "Natural language of the document (e.g., 'en', 'fr')." .
uko-doc:topicKeywords a owl:DatatypeProperty ; rdfs:domain uko:InformationUnit ; rdfs:range xsd:string ; rdfs:comment "JSON-encoded list of topic keywords extracted from this unit." .
Semantic Inference for Documents: When the document analyzer processes a document, it performs semantic analysis beyond pure structural parsing:
- Topic extraction: Each paragraph, section, and subsection is analyzed for key topics. These become
uko-doc:topicKeywordsproperties. - Implicit cross-references: If Paragraph A in Section 2 discusses "authentication flow" and Section 5 is titled "Authentication Architecture", the analyzer creates a
uko-doc:discussesTopicedge from Paragraph A to Section 5 — even though no explicit link exists in the source document. - Citation resolution: Bibliographic citations are linked to the paragraphs that cite them.
- Concept clustering: Paragraphs discussing related concepts across different sections are linked via
uko:references, enabling strategies to pull in contextually relevant content from distant parts of a document.
This semantic awareness means that a breadth-depth-navigator strategy operating on a document resource can follow not just the structural containment hierarchy but also the semantic topic graph, surfacing related sections that an author would need to consider when modifying any part of the document.
Document DetailLevelMap Refinements (per-type rendering at selected depths — see Core Data Types for the authoritative uko-doc: depth map):
| Depth | Named Level | uko-doc:Document |
uko-doc:Section |
uko-doc:Paragraph |
|---|---|---|---|---|
| 0 | TITLE_ONLY |
Document title only | Heading text only | — (not individually listed) |
| 1 | TABLE_OF_CONTENTS_L1 |
Title + top-level section headings | Heading + child section headings | First sentence only |
| 4 | TOC_WITH_SUMMARIES |
All headings + one-sentence abstract per section | Heading + paragraph count + topic keywords + abstract | First two sentences + topic keywords |
| 8 | STRUCTURAL_DETAIL |
Full TOC + figure/table captions + list item headers | Heading + first sentence of each paragraph + all captions | Full text with inline code and links preserved |
| 10 | FULL_CONTENT |
Complete document content | Complete section content including all children | Complete paragraph text with all formatting |
Data Schemas (uko-data:) — Layer 1:
The data schema ontology represents databases, data warehouses, and structured data stores. It models the schema structure (schemas, tables, columns, constraints) and captures query-time relationships (views referencing tables, foreign keys, stored procedures accessing tables). Like the document ontology, the system performs semantic analysis: when a stored procedure references columns from multiple tables, those dependency edges are created automatically.
@prefix uko-data: <https://cleveragents.ai/ontology/uko/data#> .# ── Container Classes ────────────────────────────────────────────── uko-data:Database a owl:Class ; rdfs:subClassOf uko:Container ; rdfs:comment "A database instance containing schemas." .
uko-data:Schema a owl:Class ; rdfs:subClassOf uko:Container ; rdfs:comment "A database schema (namespace for tables, views, etc.)." .
uko-data:Table a owl:Class ; rdfs:subClassOf uko:Container ; rdfs:comment "A database table containing columns." .
uko-data:View a owl:Class ; rdfs:subClassOf uko:Container ; rdfs:comment "A database view (virtual table defined by a query)." .
# ── Atom Classes ─────────────────────────────────────────────────── uko-data:Column a owl:Class ; rdfs:subClassOf uko:Atom ; rdfs:comment "A column within a table or view." .
uko-data:StoredProcedure a owl:Class ; rdfs:subClassOf uko:Atom ; rdfs:comment "A stored procedure or function in the database." .
uko-data:Trigger a owl:Class ; rdfs:subClassOf uko:Atom ; rdfs:comment "A database trigger." .
uko-data:Migration a owl:Class ; rdfs:subClassOf uko:Atom ; rdfs:comment "A schema migration (DDL change script)." .
# ── Annotation Classes ───────────────────────────────────────────── uko-data:Constraint a owl:Class ; rdfs:subClassOf uko:Annotation ; rdfs:comment "A column or table constraint (NOT NULL, UNIQUE, CHECK, etc.)." .
uko-data:Index a owl:Class ; rdfs:subClassOf uko:Annotation ; rdfs:comment "A database index on one or more columns." .
uko-data:ColumnComment a owl:Class ; rdfs:subClassOf uko:Annotation ; rdfs:comment "A COMMENT ON COLUMN annotation in the database." .
# ── Boundary Classes ─────────────────────────────────────────────── uko-data:ForeignKey a owl:Class ; rdfs:subClassOf uko:Boundary ; rdfs:comment "A foreign key constraint — an interface point between tables." .
uko-data:APIEndpoint a owl:Class ; rdfs:subClassOf uko:Boundary ; rdfs:comment "A database-level API endpoint (e.g., PostgREST, GraphQL)." .
# ── Data-Specific Relationships ──────────────────────────────────── uko-data:foreignKeyTo a owl:ObjectProperty ; rdfs:subPropertyOf uko:dependsOn ; rdfs:domain uko-data:Column ; rdfs:range uko-data:Column ; rdfs:comment "Foreign key relationship from this column to a column in another table." .
uko-data:viewReferences a owl:ObjectProperty ; rdfs:subPropertyOf uko:dependsOn ; rdfs:domain uko-data:View ; rdfs:range uko-data:Table ; rdfs:comment "This view's query references columns from the target table." .
uko-data:procedureAccesses a owl:ObjectProperty ; rdfs:subPropertyOf uko:dependsOn ; rdfs:domain uko-data:StoredProcedure ; rdfs:range uko-data:Table ; rdfs:comment "This stored procedure reads from or writes to the target table." .
uko-data:triggerFires a owl:ObjectProperty ; rdfs:subPropertyOf uko:dependsOn ; rdfs:domain uko-data:Trigger ; rdfs:range uko-data:Table ; rdfs:comment "This trigger fires on events on the target table." .
uko-data:migrationAlters a owl:ObjectProperty ; rdfs:subPropertyOf uko:dependsOn ; rdfs:domain uko-data:Migration ; rdfs:range uko-data:Table ; rdfs:comment "This migration modifies the schema of the target table." .
# ── Data-Specific Properties ────────────────────────────────────── uko-data:dataType a owl:DatatypeProperty ; rdfs:domain uko-data:Column ; rdfs:range xsd:string ; rdfs:comment "SQL data type (e.g., 'VARCHAR(255)', 'INTEGER', 'JSONB')." .
uko-data:isNullable a owl:DatatypeProperty ; rdfs:domain uko-data:Column ; rdfs:range xsd:boolean .
uko-data:isPrimaryKey a owl:DatatypeProperty ; rdfs:domain uko-data:Column ; rdfs:range xsd:boolean .
uko-data:rowEstimate a owl:DatatypeProperty ; rdfs:domain uko-data:Table ; rdfs:range xsd:long ; rdfs:comment "Estimated row count from database statistics." .
uko-data:viewDefinition a owl:DatatypeProperty ; rdfs:domain uko-data:View ; rdfs:range xsd:string ; rdfs:comment "The SQL query that defines this view." .
uko-data:procedureBody a owl:DatatypeProperty ; rdfs:domain uko-data:StoredProcedure ; rdfs:range xsd:string ; rdfs:comment "The SQL/PL body of the stored procedure." .
Database DetailLevelMap Refinements (per-type rendering at selected depths — see Core Data Types for the authoritative uko-data: depth map):
| Depth | Named Level | uko-data:Database |
uko-data:Table |
uko-data:Column |
|---|---|---|---|---|
| 0 | SCHEMA_LISTING |
Schema names only | — (not individually listed) | — |
| 1 | TABLE_LISTING |
Schema names + table/view names | Table name only | — |
| 3 | TYPED_COLUMNS |
+ column types + nullability | Column names + types + nullability | Name + type + nullable |
| 5 | RELATIONSHIPS |
+ foreign key graph + row estimates | + primary key + foreign key targets + row estimate | + FK references + constraints |
| 7 | DDL |
Full CREATE TABLE DDL with constraints | Full CREATE TABLE DDL + index definitions | Full column DDL + all constraints |
| 11 | FULL_CATALOG |
Complete DDL + views + procedures + sample data | Full DDL + triggers + sample rows + statistics | Full DDL + value distribution + sample values |
Infrastructure (uko-infra:) — Layer 1:
@prefix uko-infra: <https://cleveragents.ai/ontology/uko/infra#> .# ── Container Classes ────────────────────────────────────────────── uko-infra:Service a owl:Class ; rdfs:subClassOf uko:Container ; rdfs:comment "A deployable service or application." .
uko-infra:ConfigBlock a owl:Class ; rdfs:subClassOf uko:Container ; rdfs:comment "A configuration block or section (e.g., a YAML map, TOML table)." .
uko-infra:DeploymentUnit a owl:Class ; rdfs:subClassOf uko:Container ; rdfs:comment "A deployment unit (e.g., Kubernetes Deployment, Docker Compose service)." .
# ── Atom Classes ─────────────────────────────────────────────────── uko-infra:ConfigKey a owl:Class ; rdfs:subClassOf uko:Atom ; rdfs:comment "A configuration key-value pair." .
uko-infra:EnvironmentVariable a owl:Class ; rdfs:subClassOf uko:Atom ; rdfs:comment "An environment variable definition." .
# ── Boundary Classes ─────────────────────────────────────────────── uko-infra:Endpoint a owl:Class ; rdfs:subClassOf uko:Boundary ; rdfs:comment "A network endpoint (HTTP route, gRPC service, message queue)." .
uko-infra:Port a owl:Class ; rdfs:subClassOf uko:Boundary ; rdfs:comment "A network port binding." .
# ── Infrastructure-Specific Relationships ────────────────────────── uko-infra:connectsTo a owl:ObjectProperty ; rdfs:subPropertyOf uko:dependsOn ; rdfs:domain uko-infra:Service ; rdfs:range uko-infra:Service ; rdfs:comment "This service connects to or depends on the target service." .
uko-infra:exposes a owl:ObjectProperty ; rdfs:domain uko-infra:Service ; rdfs:range uko-infra:Endpoint ; rdfs:comment "This service exposes the target endpoint." .
Layer 2: Paradigm / Format Specializations
Layer 2 refines Layer 1 domains with paradigm- or format-specific specializations. For code: OO, functional, and procedural paradigms. For documents: academic vs. technical. For databases: relational vs. graph. Each may extend the parent domain's DetailLevelMap with additional named levels.
Object-Oriented (uko-oo:):
@prefix uko-oo: <https://cleveragents.ai/ontology/uko/oo#> .uko-oo:Class a owl:Class ; rdfs:subClassOf uko-code:TypeDefinition, uko:Container ; rdfs:comment "An OO class that contains methods and attributes." .
uko-oo:Interface a owl:Class ; rdfs:subClassOf uko-code:TypeDefinition, uko:Boundary ; rdfs:comment "An interface or abstract base class." .
uko-oo:Method a owl:Class ; rdfs:subClassOf uko-code:Callable .
uko-oo:Attribute a owl:Class ; rdfs:subClassOf uko:Atom .
uko-oo:inheritsFrom a owl:ObjectProperty ; rdfs:subPropertyOf uko:dependsOn ; rdfs:domain uko-oo:Class ; rdfs:range uko-oo:Class .
uko-oo:implements a owl:ObjectProperty ; rdfs:subPropertyOf uko:dependsOn ; rdfs:domain uko-oo:Class ; rdfs:range uko-oo:Interface .
Functional (uko-func:):
@prefix uko-func: <https://cleveragents.ai/ontology/uko/func#> .uko-func:PureFunction a owl:Class ; rdfs:subClassOf uko-code:Callable .
uko-func:TypeClass a owl:Class ; rdfs:subClassOf uko-code:TypeDefinition, uko:Boundary .
uko-func:Monad a owl:Class ; rdfs:subClassOf uko-code:TypeDefinition .
Code DetailLevelMap Refinements by Paradigm and Language:
The authoritative DetailLevelMap definitions — with full integer depth assignments and named level descriptions for each domain — are specified in Core Concepts > Advanced Context Management System (ACMS) > Core Data Types > DetailDepth and DetailLevelMap. The tables below show how paradigm-specific (Layer 2) and language-specific (Layer 3) UKO node types map their content at selected depth levels from the parent domain's map. These are illustrative excerpts showing what content each node type contributes at key depths, not exhaustive depth-by-depth definitions.
General Software (uko-code:) — Layer 1 (per-type rendering at selected depths):
| Depth | Named Level | uko-code:Module |
uko-code:Callable |
uko-code:TypeDefinition |
|---|---|---|---|---|
| 0 | MODULE_LISTING |
Module name only | — (not individually listed) | — (not individually listed) |
| 2 | MEMBER_LISTING |
+ names of top-level members | Function name (no signature) | Type name + kind (class/struct/enum) |
| 4 | SIGNATURES |
+ all top-level signatures | Name + parameter types + return type | Name + field types + method signatures |
| 6 | STRUCTURAL_OUTLINE |
+ control flow outline of callables | Signature + control flow outline (branches, loops) | Name + all field + method signatures + inheritance |
| 9 | FULL_SOURCE |
Complete module source | Complete function body | Complete type definition |
Object-Oriented (uko-oo:) — Layer 2 (per-type rendering at selected depths):
| Depth | Named Level | uko-oo:Class |
uko-oo:Method |
uko-oo:Interface |
|---|---|---|---|---|
| 2 | MEMBER_LISTING |
Class name + parent classes | Method name + visibility | Interface name |
| 3 | CLASS_HIERARCHY (OO-inserted) |
+ inheritance chain + interface implementations | — | + method signatures |
| 4 | SIGNATURES |
+ attribute types + method signatures | + params + return type + visibility | + all method signatures with param types |
| 5 | SIGNATURES_WITH_DOCS |
+ class docstring + attribute descriptions | + docstring + side effect hints | + description + implementor count |
| 9+ | FULL_SOURCE |
Complete class definition | Complete method body | Complete interface definition |
Functional (uko-func:) — Layer 2 (per-type rendering at selected depths):
| Depth | Named Level | uko-func:PureFunction |
uko-func:TypeClass |
uko-func:Monad |
|---|---|---|---|---|
| 2 | MEMBER_LISTING |
Function name | Type class name | Monad name |
| 4 | SIGNATURES |
Name + type signature + purity annotation | Name + associated types | Name + bind/return types |
| 6 | STRUCTURAL_OUTLINE |
+ pattern match structure + guard conditions | + all method signatures + laws | + bind/return implementations |
| 9 | FULL_SOURCE |
Complete function definition | Complete type class definition | Complete monad definition |
Procedural (uko-proc:) — Layer 2 (rendering at selected depths):
| Depth | Named Level | Procedural Specifics |
|---|---|---|
| 2 | MEMBER_LISTING |
Function names + global variable names (no types) |
| 4 | SIGNATURES |
Function declarations + global variable types + header file exports |
| 5 | SIGNATURES_WITH_DOCS |
+ function descriptions + global variable purposes + header summaries |
| 6 | STRUCTURAL_OUTLINE |
+ function signatures with parameter names + struct definitions + macro signatures |
| 9 | FULL_SOURCE |
Complete source including all function bodies, macros, and inline assembly |
Layer 3: Technology-Specific Specializations
Layer 3 extends Layer 2 (or Layer 1 directly) with technology-specific refinements — e.g., Python, TypeScript, Rust, Java for code; PostgreSQL, MongoDB for databases; Markdown, LaTeX for documents. Each may insert additional named DetailLevelMap levels into its parent's map.
Language-Specific (Layer 3) DetailLevelMap insertions — examples:
Each language extension may insert additional named levels into its parent paradigm's map, shifting subsequent depths upward to maintain consecutive integer numbering (see the fully expanded effective maps in Core Data Types). The table below summarizes the key additions:
| Language | Inserted Levels | Additional Content at Those Depths |
|---|---|---|
Python (uko-py:) |
DECORATED_SIGNATURES (between SIGNATURES_WITH_DOCS and STRUCTURAL_OUTLINE), TYPE_STUBS (between KEY_LOGIC and NEAR_COMPLETE), WITH_TESTS (beyond FULL_SOURCE) |
Decorator chains (@property, @staticmethod, custom), .pyi stub type annotations, associated test cases |
TypeScript (uko-ts:) |
(extends at SIGNATURES level) | export/default export markers, declare ambient types, generic type parameters, mapped/conditional types |
Rust (uko-rs:) |
(extends at SIGNATURES level) | Visibility modifiers (pub, pub(crate)), lifetime parameters, trait bounds, unsafe markers, #[derive] macros, #[must_use] annotations |
Java (uko-java:) |
(extends at SIGNATURES level) | Package visibility, final/abstract/sealed modifiers, annotation lists (@Override, @Deprecated), checked exceptions |
Not every language requires a Layer 3 extension. If a language's semantics are fully captured by the Layer 1 (uko-code:) and Layer 2 (paradigm) maps, no Layer 3 extension is needed. The system falls back to the nearest parent layer's DetailLevelMap.
The Universal View Guarantee
Because every domain-specific class ultimately inherits from Layer 0 (uko:InformationUnit, uko:Container, etc.), any strategy or query written against Layer 0 automatically works across all domains — not just different programming languages, but across code, documents, databases, and infrastructure alike. For example:
# "Find all containers that depend on this atom" — works for Python classes,
# TypeScript modules, Rust crates, SQL tables, document sections, or config blocks.
SELECT ?container WHERE {
?container a uko:Container .
?container uko:dependsOn <target_atom> .
}
Similarly, a DetailDepth request at Layer 0 is resolved by the most specific DetailLevelMap available in the inheritance chain. A request for depth 3 of a uko:Container will produce a member summary with one-line docstrings for a Python module (MEMBER_SUMMARY), a full table of contents for a document chapter (FULL_TOC), or typed column definitions for a database table (TYPED_COLUMNS) — all through the same universal integer-depth interface, with each domain mapping the integer to its domain-appropriate rendering.
Provenance Contract
Every UKO node carries provenance back to the originating resource:
uko-py:class/AuthManager
uko:sourceResource <resource-ulid-01HXM7...>
uko:sourcePath "src/auth/manager.py"
uko:sourceRange "15:1-87:0"
uko:validFrom "2026-01-15T10:30:00Z"
uko:isCurrent true
This allows the PreambleGenerator pipeline component to generate provenance summaries in the context preamble, and allows actors to trace any piece of context back to its exact source location.
Temporal Data Model and Storage Tiers
Revision-Aware RDF
Each UKO node carries temporal metadata. When code changes, existing nodes are not deleted — instead, their validUntil is set and isCurrent becomes false, and a new node is created with isRevisionOf pointing to the predecessor:
# Before code change: uko-py:class/AuthManager_v1 uko:isCurrent true ; uko:validFrom "2026-01-10T00:00:00Z"^^xsd:dateTime .# After code change: uko-py:class/AuthManager_v1 uko:isCurrent false ; uko:validFrom "2026-01-10T00:00:00Z"^^xsd:dateTime ; uko:validUntil "2026-01-15T10:30:00Z"^^xsd:dateTime .
uko-py:class/AuthManager_v2 uko:isCurrent true ; uko:validFrom "2026-01-15T10:30:00Z"^^xsd:dateTime ; uko:isRevisionOf uko-py:class/AuthManager_v1 .
This temporal chain is what enables the temporal-archaeology strategy to find patterns like "this class was refactored 3 times in the last month" or "this function's signature changed after we updated the auth library."
Three Storage Tiers with Temporal Alignment
| Tier | Content | Retention | Access Pattern | Temporal |
|---|---|---|---|---|
| Hot | Current UKO graph + recent embeddings + active text index | Until resource removed | Direct query via BAL | isCurrent = true only |
| Warm | Recent decision contexts, plan context snapshots | context.tiers.warm.retention-hours (default: 24h) |
Scoped query via plan hierarchy | Current + recently-expired nodes |
| Cold | Archived decision contexts, historical UKO snapshots | context.tiers.cold.retention-days (default: 90d) |
Full historical query | All temporal versions |
Scoped Views and Plan Subgraph Projection
Resource Scope Resolution
When a plan is created (via agents plan use), the system resolves its resource scope:
class ResourceScopeResolver: """Determines which resources a plan can see."""<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">resolve</span>(<span style="color: cyan;">self</span>, plan: <span style="color: cyan;">Plan</span>) -> <span style="color: cyan;">ResourceScope</span>: <span style="color: #888;"># 1. Get all projects this plan targets</span> projects = plan.target_projects <span style="color: #888;"># 2. Collect all resources linked to those projects</span> resources = <span style="color: cyan;">set</span>() <span style="color: magenta; font-weight: 600;">for</span> project <span style="color: magenta; font-weight: 600;">in</span> projects: resources.update(project.linked_resources) <span style="color: #888;"># 3. Expand to include child resources (DAG traversal)</span> expanded = <span style="color: cyan;">set</span>() <span style="color: magenta; font-weight: 600;">for</span> resource <span style="color: magenta; font-weight: 600;">in</span> resources: expanded.add(resource) expanded.update(<span style="color: cyan;">self</span>.registry.get_descendants(resource)) <span style="color: #888;"># 4. Apply context view filters (include/exclude)</span> view = <span style="color: cyan;">self</span>._resolve_view(plan) filtered = <span style="color: cyan;">self</span>._apply_filters(expanded, view) <span style="color: magenta; font-weight: 600;">return</span> ResourceScope(resources=filtered, projects=projects)
UKO Subgraph Projection
Given a resource scope, the system projects only the relevant UKO subgraph:
class UKOSubgraphProjector: """Projects the UKO graph to only include nodes from scoped resources."""<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">project</span>(<span style="color: cyan;">self</span>, scope: <span style="color: cyan;">ResourceScope</span>, graph_backend: <span style="color: cyan;">GraphBackend</span>) -> <span style="color: cyan;">ProjectedGraph</span>: <span style="color: #888;"># Only include UKO nodes whose sourceResource is in scope</span> <span style="color: magenta; font-weight: 600;">return</span> graph_backend.subgraph_by_resources( resource_ulids=[r.ulid <span style="color: magenta; font-weight: 600;">for</span> r <span style="color: magenta; font-weight: 600;">in</span> scope.resources], include_temporal=scope.temporal_scope, )
Scoped Backend Views
Each backend (text, vector, graph) is wrapped in a ScopedView that automatically filters queries to only return results from in-scope resources:
class ScopedBackendView: """Wraps a backend to restrict queries to a resource scope."""<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">__init__</span>(<span style="color: cyan;">self</span>, backend, scope: <span style="color: cyan;">ResourceScope</span>): <span style="color: cyan;">self</span>.backend = backend <span style="color: cyan;">self</span>.scope = scope <span style="color: cyan;">self</span>._resource_filter = {r.ulid <span style="color: magenta; font-weight: 600;">for</span> r <span style="color: magenta; font-weight: 600;">in</span> scope.resources} <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">search</span>(<span style="color: cyan;">self</span>, query, **kwargs): <span style="color: #888;"># Inject the resource filter into the backend query</span> results = <span style="color: cyan;">self</span>.backend.search( query, resource_filter=<span style="color: cyan;">self</span>._resource_filter, **kwargs, ) <span style="color: magenta; font-weight: 600;">return</span> results
This is critical for multi-project isolation: a plan targeting local/api-service never sees resources from local/frontend, even if both projects share the same physical database indices.
Backend Abstraction Layer (BAL)
The BAL provides a uniform interface to heterogeneous data stores. Each backend type has a protocol:
Backend Protocols
class TextBackend(Protocol): def search(self, query: str, *, resource_filter: set[str], limit: int = 20) -> list[TextResult]: ... def index(self, resource_ulid: str, content: str, metadata: dict) -> None: ...class VectorBackend(Protocol): def search(self, embedding: list[float], *, resource_filter: set[str], limit: int = 20, threshold: float = 0.3) -> list[VectorResult]: ... def index(self, resource_ulid: str, chunks: list[EmbeddingChunk]) -> None: ...
class GraphBackend(Protocol): def query_sparql(self, sparql: str, *, resource_filter: set[str]) -> list[dict]: ... def add_triples(self, triples: list[Triple]) -> None: ... def subgraph_by_resources(self, resource_ulids: list[str], include_temporal: TemporalScope) -> ProjectedGraph: ... def traverse(self, start_uri: str, max_hops: int, edge_types: list[str], resource_filter: set[str]) -> SubGraph: ...
Unified Result Types
@dataclass class TextResult: resource_ulid: str path: str line_range: tuple[int, int] content: str score: float
@dataclass class VectorResult: resource_ulid: str uko_node: str content: str similarity: float metadata: dict
Context Assembly Pipeline
The context assembly pipeline is the central orchestrator of context assembly. It receives a ContextRequest, processes it through a chain of pluggable components, and produces a budget-respecting AssembledContext. Every stage of the pipeline is a replaceable plugin — each defined by a Protocol interface, shipped with a default implementation, and configurable at the global, project, or plan scope.
Pipeline Component Model
The pipeline follows the Pipes and Filters architectural pattern, with each filter implemented as a Strategy (GoF) that can be swapped at runtime. Component resolution uses a Chain of Responsibility pattern: plan-level overrides take precedence over project-level, which take precedence over global defaults.
Pipeline Component Resolution Order:
plan scope ──► project scope ──► global scope ──► built-in default (most specific) (least specific)
There are ten pluggable components, grouped into three pipeline phases:
Phase 1 — Strategy Orchestration (selects and runs context strategies):
| # | Component | Protocol | Responsibility |
|---|---|---|---|
| 1 | StrategySelector | StrategySelectorProtocol |
Decides which strategies to invoke and computes confidence scores |
| 2 | BudgetAllocator | BudgetAllocatorProtocol |
Distributes the token budget across selected strategies |
| 3 | StrategyExecutor | StrategyExecutorProtocol |
Controls how strategies are invoked (parallelism, timeouts, error handling) |
Phase 2 — Fragment Fusion (merges strategy outputs into a coherent payload):
| # | Component | Protocol | Responsibility |
|---|---|---|---|
| 4 | FragmentDeduplicator | FragmentDeduplicatorProtocol |
Removes duplicate fragments across strategy outputs |
| 5 | DetailDepthResolver | DetailDepthResolverProtocol |
Resolves conflicts when the same UKO node appears at different depths |
| 6 | FragmentScorer | FragmentScorerProtocol |
Computes composite relevance scores for ranking |
| 7 | BudgetPacker | BudgetPackerProtocol |
Fits scored fragments into the token budget |
| 8 | FragmentOrderer | FragmentOrdererProtocol |
Orders final fragments for coherence in the prompt |
Phase 3 — Context Finalization (produces the final deliverable):
| # | Component | Protocol | Responsibility |
|---|---|---|---|
| 9 | PreambleGenerator | PreambleGeneratorProtocol |
Generates the context preamble (provenance summary, structure map) |
| 10 | SkeletonCompressor | SkeletonCompressorProtocol |
Compresses parent context into a skeleton for child plan inheritance |
Component Protocol Definitions
Each component is defined as a @runtime_checkable Protocol, enabling both structural subtyping (duck typing) and explicit isinstance checks. The protocols use Generic type parameters where applicable, and employ the Template Method pattern — each protocol defines the contract while default implementations provide the algorithm skeleton with overridable hook methods.
# ── Phase 1: Strategy Orchestration ────────────────────────────────@runtime_checkable class StrategySelectorProtocol(Protocol): """Decides which strategies to invoke and with what confidence. Applies request preferences and backend availability filtering."""
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">select</span>( <span style="color: cyan;">self</span>, strategies: <span style="color: cyan;">Sequence</span>[ContextStrategy], request: ContextRequest, backends: BackendSet, ) -> <span style="color: cyan;">list</span>[<span style="color: cyan;">tuple</span>[ContextStrategy, <span style="color: cyan;">float</span>]]: <span style="color: #888;">"""Returns (strategy, confidence) pairs, sorted by priority. Confidence is 0.0-1.0. Strategies with 0.0 are excluded."""</span> ...@runtime_checkable class BudgetAllocatorProtocol(Protocol): """Distributes the token budget across selected strategies. May use proportional, priority-weighted, or custom allocation schemes."""
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">allocate</span>( <span style="color: cyan;">self</span>, candidates: <span style="color: cyan;">list</span>[<span style="color: cyan;">tuple</span>[ContextStrategy, <span style="color: cyan;">float</span>]], total_budget: <span style="color: cyan;">int</span>, request: ContextRequest, ) -> <span style="color: cyan;">list</span>[<span style="color: cyan;">tuple</span>[ContextStrategy, <span style="color: cyan;">float</span>, <span style="color: cyan;">int</span>]]: <span style="color: #888;">"""Returns (strategy, confidence, allocated_tokens) triples. Sum of allocated_tokens must not exceed total_budget."""</span> ...@runtime_checkable class StrategyExecutorProtocol(Protocol): """Controls how strategies are invoked — parallelism model, timeout handling, circuit breaking, and error recovery."""
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">execute</span>( <span style="color: cyan;">self</span>, allocations: <span style="color: cyan;">list</span>[<span style="color: cyan;">tuple</span>[ContextStrategy, <span style="color: cyan;">float</span>, <span style="color: cyan;">int</span>]], request: ContextRequest, backends: BackendSet, plan_context: PlanContext, ) -> <span style="color: cyan;">list</span>[ContextFragment]: <span style="color: #888;">"""Executes strategies and collects all fragments. Must handle strategy failures gracefully (log and continue)."""</span> ...# ── Phase 2: Fragment Fusion ────────────────────────────────────────
@runtime_checkable class FragmentDeduplicatorProtocol(Protocol): """Removes duplicate fragments from the merged strategy output. Deduplication can be identity-based, content-hash-based, or semantic."""
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">deduplicate</span>( <span style="color: cyan;">self</span>, fragments: <span style="color: cyan;">list</span>[ContextFragment], ) -> <span style="color: cyan;">list</span>[ContextFragment]: <span style="color: #888;">"""Returns deduplicated fragments. When duplicates are found, the fragment with the highest relevance_score is retained."""</span> ...@runtime_checkable class DetailDepthResolverProtocol(Protocol): """Resolves conflicts when the same UKO node appears at different detail depths across strategy outputs."""
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">resolve</span>( <span style="color: cyan;">self</span>, fragments: <span style="color: cyan;">list</span>[ContextFragment], budget: <span style="color: cyan;">int</span>, ) -> <span style="color: cyan;">list</span>[ContextFragment]: <span style="color: #888;">"""Returns fragments with depth conflicts resolved. Default: keep highest depth that fits within budget."""</span> ...@runtime_checkable class FragmentScorerProtocol(Protocol): """Computes a composite score for each fragment, used for ranking during budget packing. Applies hierarchical weighting, strategy quality weighting, and recency bonuses."""
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">score</span>( <span style="color: cyan;">self</span>, fragments: <span style="color: cyan;">list</span>[ContextFragment], plan_context: PlanContext, ) -> <span style="color: cyan;">list</span>[ScoredFragment]: <span style="color: #888;">"""Returns fragments annotated with composite scores. ScoredFragment adds a `composite_score: float` field."""</span> ...@runtime_checkable class BudgetPackerProtocol(Protocol): """Fits scored fragments into the token budget using a packing algorithm. Supports depth fallback — if a fragment doesn't fit at its current depth, it can be re-rendered at a lower depth."""
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">pack</span>( <span style="color: cyan;">self</span>, scored_fragments: <span style="color: cyan;">list</span>[ScoredFragment], budget: <span style="color: cyan;">int</span>, detail_level_maps: DetailLevelMapRegistry, ) -> <span style="color: cyan;">list</span>[ContextFragment]: <span style="color: #888;">"""Returns the subset of fragments that fit within budget, possibly re-rendered at lower depths. Ordered by score."""</span> ...@runtime_checkable class FragmentOrdererProtocol(Protocol): """Orders the packed fragments for optimal coherence in the actor's context window. Ordering may be by relevance, by topological dependency, or by a hybrid coherence heuristic."""
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">order</span>( <span style="color: cyan;">self</span>, fragments: <span style="color: cyan;">list</span>[ContextFragment], ) -> <span style="color: cyan;">list</span>[ContextFragment]: <span style="color: #888;">"""Returns fragments in the final presentation order."""</span> ...# ── Phase 3: Context Finalization ──────────────────────────────────
@runtime_checkable class PreambleGeneratorProtocol(Protocol): """Generates the preamble prepended to assembled context — a structured summary of what's included, why, and from where."""
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">generate</span>( <span style="color: cyan;">self</span>, fragments: <span style="color: cyan;">list</span>[ContextFragment], strategies_used: <span style="color: cyan;">list</span>[<span style="color: cyan;">str</span>], budget_used: <span style="color: cyan;">float</span>, max_tokens: <span style="color: cyan;">int</span>, ) -> <span style="color: cyan;">str</span> | <span style="color: magenta; font-weight: 600;">None</span>: <span style="color: #888;">"""Returns the preamble string, or None if preamble is disabled."""</span> ...@runtime_checkable class SkeletonCompressorProtocol(Protocol): """Compresses a parent plan's assembled context into a compact skeleton representation for inheritance by child plans."""
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">compress</span>( <span style="color: cyan;">self</span>, parent_context: AssembledContext, child_focus: <span style="color: cyan;">list</span>[<span style="color: cyan;">str</span>], skeleton_budget: <span style="color: cyan;">int</span>, ) -> AssembledContext: <span style="color: #888;">"""Returns a compressed version of the parent context that fits within skeleton_budget tokens. Typically reduces all fragments to depth 0-1 (e.g., MODULE_LISTING / TITLE_ONLY)."""</span> ...
Supporting Types
@dataclass(frozen=True) class ScoredFragment: """A ContextFragment annotated with a composite score by the FragmentScorer.""" fragment: ContextFragment composite_score: float # 0.0-1.0 composite ranking score score_components: dict[str, float] # Breakdown: {"relevance": 0.9, "hierarchy": 0.8, ...}@dataclass class PipelineConfig: """Resolved configuration for the context assembly pipeline. Each field holds the concrete plugin instance to use for that stage. Resolved via the scope chain: plan > project > global > built-in."""
strategy_selector: StrategySelectorProtocol budget_allocator: BudgetAllocatorProtocol strategy_executor: StrategyExecutorProtocol fragment_deduplicator: FragmentDeduplicatorProtocol detail_depth_resolver: DetailDepthResolverProtocol fragment_scorer: FragmentScorerProtocol budget_packer: BudgetPackerProtocol fragment_orderer: FragmentOrdererProtocol preamble_generator: PreambleGeneratorProtocol skeleton_compressor: SkeletonCompressorProtocol
Pipeline Orchestrator
The ContextAssemblyPipeline is the top-level orchestrator that wires the ten components together. It implements the Mediator pattern — components do not communicate directly with each other; the pipeline mediates all data flow. The pipeline itself is stateless; all state is carried through the data flowing between stages.
class ContextAssemblyPipeline: """Mediator that orchestrates the 10-stage context assembly pipeline. Each stage is a pluggable component resolved from the scope chain."""<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">__init__</span>(<span style="color: cyan;">self</span>, config: PipelineConfig, strategies: <span style="color: cyan;">Sequence</span>[ContextStrategy]): <span style="color: cyan;">self</span>._config = config <span style="color: cyan;">self</span>._strategies = strategies <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">assemble</span>( <span style="color: cyan;">self</span>, request: ContextRequest, backends: BackendSet, budget: <span style="color: cyan;">int</span>, plan_context: PlanContext, ) -> AssembledContext: <span style="color: #888;"># ── Phase 1: Strategy Orchestration ──────────────────────</span> <span style="color: #888;"># 1. Select strategies</span> candidates = <span style="color: cyan;">self</span>._config.strategy_selector.select( <span style="color: cyan;">self</span>._strategies, request, backends, ) <span style="color: #888;"># 2. Allocate budget</span> allocations = <span style="color: cyan;">self</span>._config.budget_allocator.allocate( candidates, budget, request, ) <span style="color: #888;"># 3. Execute strategies</span> raw_fragments = <span style="color: cyan;">self</span>._config.strategy_executor.execute( allocations, request, backends, plan_context, ) <span style="color: #888;"># ── Phase 2: Fragment Fusion ──────────────────────────────</span> <span style="color: #888;"># 4. Deduplicate</span> deduped = <span style="color: cyan;">self</span>._config.fragment_deduplicator.deduplicate(raw_fragments) <span style="color: #888;"># 5. Resolve depth conflicts</span> resolved = <span style="color: cyan;">self</span>._config.detail_depth_resolver.resolve(deduped, budget) <span style="color: #888;"># 6. Score fragments</span> scored = <span style="color: cyan;">self</span>._config.fragment_scorer.score(resolved, plan_context) <span style="color: #888;"># 7. Pack into budget</span> packed = <span style="color: cyan;">self</span>._config.budget_packer.pack( scored, budget, <span style="color: cyan;">self</span>._detail_level_maps, ) <span style="color: #888;"># 8. Order for coherence</span> ordered = <span style="color: cyan;">self</span>._config.fragment_orderer.order(packed) <span style="color: #888;"># ── Phase 3: Context Finalization ─────────────────────────</span> <span style="color: #888;"># 9. Generate preamble</span> strategies_used = [s.name <span style="color: magenta; font-weight: 600;">for</span> s, _, _ <span style="color: magenta; font-weight: 600;">in</span> allocations] preamble = <span style="color: cyan;">self</span>._config.preamble_generator.generate( ordered, strategies_used, budget_used=<span style="color: cyan;">sum</span>(f.token_count <span style="color: magenta; font-weight: 600;">for</span> f <span style="color: magenta; font-weight: 600;">in</span> ordered) / budget, max_tokens=<span style="color: cyan;">self</span>._preamble_max_tokens, ) <span style="color: magenta; font-weight: 600;">return</span> AssembledContext( fragments=ordered, total_tokens=<span style="color: cyan;">sum</span>(f.token_count <span style="color: magenta; font-weight: 600;">for</span> f <span style="color: magenta; font-weight: 600;">in</span> ordered), budget_used=<span style="color: cyan;">sum</span>(f.token_count <span style="color: magenta; font-weight: 600;">for</span> f <span style="color: magenta; font-weight: 600;">in</span> ordered) / budget, strategies_used=strategies_used, context_hash=<span style="color: cyan;">self</span>._compute_hash(ordered), preamble=preamble, provenance_map={f.uko_node: f.provenance <span style="color: magenta; font-weight: 600;">for</span> f <span style="color: magenta; font-weight: 600;">in</span> ordered}, )
Built-in Default Implementations
Each component ships with a default implementation that provides the behavior described in the current specification. Custom implementations need only implement the relevant Protocol — they need not subclass the default.
Phase 1 defaults:
| Component | Default Class | Algorithm |
|---|---|---|
| StrategySelector | ConfidenceWeightedSelector |
Calls can_handle() on all registered strategies; filters to confidence > 0; applies ContextRequest.preferred_strategies preferences; sorts by confidence * quality_score descending. Uses the Observer pattern to emit StrategySelectionEvent for diagnostics. |
| BudgetAllocator | ProportionalBudgetAllocator |
Allocates budget proportional to confidence * quality_score. Guarantees each strategy receives at least min_useful_budget tokens or is excluded. Implements Flyweight for allocation metadata reuse across re-assembly cycles. |
| StrategyExecutor | ParallelStrategyExecutor |
Runs strategies concurrently via ThreadPoolExecutor. Per-strategy timeout (default: 30s). Uses Circuit Breaker pattern: after 3 consecutive failures, a strategy is temporarily disabled for that plan. Failed strategies log warnings and are excluded from results. |
Phase 2 defaults:
| Component | Default Class | Algorithm |
|---|---|---|
| FragmentDeduplicator | ContentHashDeduplicator |
Groups fragments by UKO node URI; within each group, hashes content to detect duplicates; retains the fragment with the highest relevance_score. Configurable via context.fusion.dedup-strategy: content-hash (default), uko-identity (URI only), or semantic (embedding cosine similarity > 0.95). Implements Strategy (GoF) pattern to swap dedup algorithms. |
| DetailDepthResolver | MaxDepthResolver |
When the same UKO node appears at multiple depths, retains the highest-depth rendering that fits within the remaining budget. Uses a Specification pattern: each fragment is tested against a budget specification, and the highest-depth satisfying fragment wins. |
| FragmentScorer | WeightedCompositeScorer |
Computes composite = relevance_score * strategy_quality * hierarchy_weight * recency_bonus. Hierarchy weight is derived from the fragment's distance to the plan's focus nodes. Recency bonus is 1.0 for current nodes, decaying for older revisions. Scoring weights are configurable. Uses Decorator pattern to compose scoring dimensions. |
| BudgetPacker | GreedyKnapsackPacker |
Sorts by composite score descending; greedily adds fragments until budget is exhausted. When a fragment doesn't fit at its current depth, attempts depth fallback: re-renders at progressively lower depths (e.g., depth 9 → 4 → 0) until it fits or is excluded. Implements Iterator pattern over the depth fallback sequence. |
| FragmentOrderer | RelevanceCoherenceOrderer |
Groups fragments by resource, then orders within each group by containment depth and source position. Groups are ordered by maximum composite score within the group. Configurable via context.fusion.ordering: relevance (strict score order), topological (dependency-aware), relevance-coherence (default hybrid). Implements Template Method with overridable _group_key() and _intra_group_sort() hooks. |
Phase 3 defaults:
| Component | Default Class | Algorithm |
|---|---|---|
| PreambleGenerator | ProvenancePreambleGenerator |
Generates a structured preamble listing: included resources, strategies used, budget utilization, and a compact structure map. Respects context.fusion.preamble-max-tokens limit. Can be disabled via context.fusion.preamble-enabled = false. Uses Builder pattern to construct preamble sections. |
| SkeletonCompressor | DepthReductionCompressor |
Re-renders all parent context fragments at depth 0-1 (e.g., MODULE_LISTING for code, TABLE_OF_CONTENTS_L1 for documents). Prioritizes fragments closer to the child's focus area. Fits within skeleton_budget tokens (derived from skeleton_ratio). Uses Visitor pattern to traverse and re-render fragments by domain type. |
Scope Resolution and Configuration
Pipeline components are resolved through a three-level scope chain. Each scope can override any subset of the ten components; unspecified components inherit from the next broader scope.
class PipelineConfigResolver: """Resolves the effective PipelineConfig for a given context assembly. Implements the Chain of Responsibility pattern across three scopes."""<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">resolve</span>( <span style="color: cyan;">self</span>, plan: Plan | <span style="color: magenta; font-weight: 600;">None</span>, project: Project | <span style="color: magenta; font-weight: 600;">None</span>, global_config: GlobalConfig, ) -> PipelineConfig: <span style="color: #888;"># 1. Start with built-in defaults</span> config = <span style="color: cyan;">self</span>._builtin_defaults() <span style="color: #888;"># 2. Apply global overrides from config.toml</span> config = <span style="color: cyan;">self</span>._apply_scope(config, global_config.pipeline_overrides) <span style="color: #888;"># 3. Apply project-level overrides (if any)</span> <span style="color: magenta; font-weight: 600;">if</span> project <span style="color: magenta; font-weight: 600;">and</span> project.pipeline_overrides: config = <span style="color: cyan;">self</span>._apply_scope(config, project.pipeline_overrides) <span style="color: #888;"># 4. Apply plan-level overrides (if any)</span> <span style="color: magenta; font-weight: 600;">if</span> plan <span style="color: magenta; font-weight: 600;">and</span> plan.pipeline_overrides: config = <span style="color: cyan;">self</span>._apply_scope(config, plan.pipeline_overrides) <span style="color: magenta; font-weight: 600;">return</span> config <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">_apply_scope</span>(<span style="color: cyan;">self</span>, base: PipelineConfig, overrides: <span style="color: cyan;">dict</span>) -> PipelineConfig: <span style="color: #888;">"""Merges overrides into the base config using the Prototype pattern. Only non-None overrides replace the base component."""</span> <span style="color: magenta; font-weight: 600;">return</span> PipelineConfig(**{ field: overrides.get(field, getattr(base, field)) <span style="color: magenta; font-weight: 600;">for</span> field <span style="color: magenta; font-weight: 600;">in</span> PipelineConfig.__dataclass_fields__ })
Pipeline components are registered in TOML configuration:
# ── Global config.toml ──────────────────────────────────────────── [context.pipeline] # Override any pipeline component globally (values are "module:ClassName") strategy-selector = "builtin:ConfidenceWeightedSelector" # default budget-allocator = "builtin:ProportionalBudgetAllocator" # default strategy-executor = "builtin:ParallelStrategyExecutor" # default fragment-deduplicator = "builtin:ContentHashDeduplicator" # default detail-depth-resolver = "builtin:MaxDepthResolver" # default fragment-scorer = "builtin:WeightedCompositeScorer" # default budget-packer = "builtin:GreedyKnapsackPacker" # default fragment-orderer = "builtin:RelevanceCoherenceOrderer" # default preamble-generator = "builtin:ProvenancePreambleGenerator" # default skeleton-compressor = "builtin:DepthReductionCompressor" # default# ── Per-component configuration ──────────────────────────────────── [context.pipeline.strategy-executor] timeout-seconds = 30 max-workers = 4 circuit-breaker-threshold = 3
[context.pipeline.fragment-scorer] relevance-weight = 0.4 hierarchy-weight = 0.3 quality-weight = 0.2 recency-weight = 0.1
[context.pipeline.budget-packer] depth-fallback-steps = [9, 4, 2, 0] # depths to try during fallback min-fragment-tokens = 10 # fragments below this are excluded
Project-level and plan-level overrides use the same key structure, specified via the context view YAML or the agents project context set CLI:
# ── Project-level context view YAML ────────────────────────────────
project: local/api-service
view: strategize
pipeline:
fragment-scorer: "my_extensions.scorers:DomainAwareScorer"
budget-packer: "my_extensions.packers:PriorityPreservingPacker"
Fallback Degradation Path
The pipeline's ConfidenceWeightedSelector (default StrategySelector) implements automatic fallback:
- Try
arce(requires all backends) — if unavailable: - Try
breadth-depth-navigator(requires graph) — if unavailable: - Try
semantic-embedding(requires vector) — if unavailable: - Fall back to
simple-keyword(requires only text search / ripgrep)
This ensures the system always produces context, even when advanced backends are not configured.
Initial Context Assembly
Before the actor's first turn, the system assembles initial context:
class InitialContextAssembler: """Produces the starting context for an actor invocation."""<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">assemble</span>(<span style="color: cyan;">self</span>, plan: <span style="color: cyan;">Plan</span>, actor: <span style="color: cyan;">Actor</span>, token_budget: <span style="color: cyan;">int</span>) -> <span style="color: cyan;">AssembledContext</span>: <span style="color: #888;"># 1. Resolve resource scope</span> scope = <span style="color: cyan;">self</span>.scope_resolver.resolve(plan) <span style="color: #888;"># 2. Get inherited context from parent plan (if any)</span> inherited = <span style="color: cyan;">self</span>._get_inherited_context(plan) <span style="color: #888;"># 3. Create scoped backend views</span> backends = <span style="color: cyan;">self</span>.bal.create_scoped_views(scope) <span style="color: #888;"># 4. Build the initial context request from plan metadata</span> initial_request = <span style="color: cyan;">self</span>._build_initial_request(plan, actor, inherited) <span style="color: #888;"># 5. Run through the Context Assembly Pipeline</span> result = <span style="color: cyan;">self</span>.pipeline.assemble( request=initial_request, backends=backends, budget=token_budget, plan_context=PlanContext( plan=plan, parent_context=inherited, depth_in_tree=plan.depth_in_tree, ), ) <span style="color: magenta; font-weight: 600;">return</span> result
Context Inheritance Mechanism
class PlanContextInheritance: """Manages how child plans inherit and refine parent context. Uses the pipeline's SkeletonCompressor component to produce compact parent context for child plan inheritance."""<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">compute_child_context</span>(<span style="color: cyan;">self</span>, parent_plan, child_plan, parent_context): <span style="color: #888;"># 1. Determine the child's focus from the parent's decisions</span> child_focus = <span style="color: cyan;">self</span>._extract_child_focus(parent_plan, child_plan) <span style="color: #888;"># 2. Compute depth/breadth adjustments</span> depth_delta = child_plan.depth_in_tree - parent_plan.depth_in_tree child_detail = <span style="color: cyan;">self</span>._increase_detail(<span style="color: cyan;">self</span>._avg_detail(parent_context), depth_delta) child_breadth = max(<span style="color: yellow;">1</span>, parent_context.avg_breadth - depth_delta) <span style="color: #888;"># 3. Compute skeleton budget from the child's token budget</span> skeleton_budget = <span style="color: cyan;">int</span>(child_plan.token_budget * <span style="color: cyan;">self</span>.config.skeleton_ratio) <span style="color: #888;"># 4. Use the pipeline's SkeletonCompressor to build the parent skeleton</span> skeleton = <span style="color: cyan;">self</span>.pipeline.skeleton_compressor.compress( parent_context=parent_context, child_focus=child_focus, skeleton_budget=skeleton_budget, ) <span style="color: #888;"># 5. Build the inherited context request with parent skeleton</span> <span style="color: magenta; font-weight: 600;">return</span> ContextRequest( focus=child_focus, breadth=child_breadth, depth=child_detail, depth_gradient=<span style="color: magenta; font-weight: 600;">True</span>, metadata={ <span style="color: #66cc66;">"parent_context_hash"</span>: parent_context.context_hash, <span style="color: #66cc66;">"parent_skeleton"</span>: skeleton, <span style="color: #66cc66;">"inherited_decisions"</span>: <span style="color: cyan;">self</span>._relevant_parent_decisions(parent_plan, child_focus), }, )
Skeleton Context Propagation
The skeleton is a compact, low-depth representation (depth 0-1) of the parent's context window that is passed to every child. This ensures children always have the "big picture" available, even though they focus on a narrow area. The skeleton typically consumes 5-15% of the child's token budget.
Parent's hot context (32K tokens): ┌─ Module: src/auth/ [depth 3/MEMBER_SUMMARY: 2K tokens] ├─ Module: src/core/ [depth 3/MEMBER_SUMMARY: 2K tokens] ├─ Module: src/api/ [depth 3/MEMBER_SUMMARY: 2K tokens] ├─ Module: src/services/ [depth 3/MEMBER_SUMMARY: 2K tokens] └─ ... 12 more modules [depth 3/MEMBER_SUMMARY: 24K tokens]
Child's hot context (32K tokens): ┌─ [SKELETON from parent] [depth 0/MODULE_LISTING: 3K tokens] │ Module: src/auth/ [just name + member names] │ Module: src/core/ [just name + member names] │ Module: src/api/ [just name + member names] │ ... etc. ├─ [CHILD FOCUS AREA] [depth 9/FULL_SOURCE: 25K tokens] │ Class: AuthManager [depth 9: all methods with bodies] │ Class: TokenValidator [depth 9: referenced by AuthManager] │ Class: User [depth 4/SIGNATURES: used as parameter type] └─ [INHERITED DECISIONS] [4K tokens] Decision: "Use async patterns for all auth operations" Decision: "Maintain backward compat with sync callers"
Dynamic Context Window Adaptation
The context window size may change between invocations. Different LLM providers, different models, different configurations, and even dynamic adjustments during a session can alter the available budget. The ACMS handles this seamlessly.
Budget Computation
class ContextBudgetCalculator: """Computes the effective context budget for each assembly cycle."""<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">compute</span>(<span style="color: cyan;">self</span>, actor: <span style="color: cyan;">Actor</span>, plan: <span style="color: cyan;">Plan</span>, conversation_history_tokens: <span style="color: cyan;">int</span>) -> <span style="color: cyan;">int</span>: <span style="color: #888;"># 1. Get the actor's hard context window limit</span> model_limit = actor.model_config.context_window_tokens <span style="color: #888;"># 2. Reserve space for system prompt, history, tools, response</span> reserved = (<span style="color: cyan;">self</span>._count_tokens(actor.system_prompt) + conversation_history_tokens + <span style="color: cyan;">self</span>._estimate_tool_tokens(actor.skills) + actor.model_config.max_output_tokens) <span style="color: #888;"># 3. Apply the soft cap from context view policy</span> soft_cap = <span style="color: cyan;">self</span>._resolve_view(plan).hot_max_tokens <span style="color: magenta; font-weight: 600;">or</span> <span style="color: cyan;">float</span>(<span style="color: #66cc66;">'inf'</span>) <span style="color: #888;"># 4. Calculate available budget</span> effective_budget = max(<span style="color: yellow;">0</span>, min(model_limit - reserved, soft_cap)) <span style="color: magenta; font-weight: 600;">return</span> effective_budget
Adaptive Behavior
When the budget changes between context assembly cycles:
- Budget increase: The pipeline re-runs strategy orchestration (Phase 1) with the expanded budget. The BudgetAllocator distributes the larger budget across strategies, and the BudgetPacker may promote fragments previously rendered at a low depth (e.g., depth 0) to a higher depth (e.g., depth 4 or 9).
- Budget decrease: The pipeline's BudgetPacker re-applies the knapsack with the new budget, potentially reducing fragment depths via its depth fallback mechanism or dropping low-relevance items as ranked by the FragmentScorer.
- Strategies are stateless: Each assembly cycle is independent. The pipeline runs all 10 components from scratch, and strategies receive the current budget and must respect it.
Progressive Refinement Within a Session
As an actor makes tool calls during a session, the conversation grows and the available budget shrinks. The ACMS handles this through a context refresh cycle that triggers re-assembly when the budget changes by more than 30% (configurable via context.budget.refresh_threshold).
Built-in Strategy Catalogue
The ACMS ships with several built-in strategies. The ARCE pipeline is one of them, not the only one.
Strategy: simple-keyword
Basic keyword/regex text search. Works with any backend, any resource type. No graph or vector required. This is the universal fallback. Quality score: 0.3.
Strategy: semantic-embedding
Vector similarity search. Finds semantically related content even without exact keyword matches. Requires a vector backend. Quality score: 0.6.
Strategy: breadth-depth-navigator
The graph-aware strategy that uses the depth/breadth projection system. Works with the UKO graph to provide structurally-aware context at any detail depth. Supports focus items, hop traversal, and detail depth gradients. This is the primary strategy for code-aware context. Quality score: 0.85.
Strategy: arce (Analyze, Retrieve, Contextualize, Execute)
The sophisticated multi-modal pipeline. Combines text, vector, and graph search with intent analysis and actor-type-aware patterns. This is the "kitchen sink" strategy that produces the highest quality results but requires all backends. Quality score: 0.95.
Actor-type-specific graph traversal patterns:
- Strategy actors: Broader, shallower traversal. Emphasis on module boundaries and dependency structures. Prioritizes
uko:Containeranduko:Boundarynodes. - Execution actors: Narrower, deeper traversal. Emphasis on implementation details, function bodies, and test coverage. Prioritizes
uko:Atomnodes. - Estimation actors: Focus on size metrics, complexity indicators, and historical data. Prioritizes temporal data.
Strategy: temporal-archaeology
Searches the cold tier for historical patterns. Useful for understanding "what changed last time we touched this area" or "what decisions were made about this module in the past." Requires graph + cold tier access. Quality score: 0.5.
Strategy: plan-decision-context
Retrieves context from parent and ancestor plan decisions. This is how child plans "remember" what their parent decided and why. Operates on the warm/cold tiers. Quality score: 0.7.
ACMS Extensibility
The ACMS-specific extension points (analyzers, backends, UKO vocabularies, strategies, and pipeline components) are documented in the Extensibility section below. See Extensibility > ACMS Extensions for the full details.
Real-time Index Synchronization
The system maintains index freshness through immediate, proactive updates. The UKOIndexer produces UKO triples from resources using pluggable analyzers, and simultaneously indexes into text and vector backends:
class UKOIndexer: """Produces UKO triples from resources using pluggable analyzers."""<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">index_resource</span>(<span style="color: cyan;">self</span>, resource: <span style="color: cyan;">Resource</span>): <span style="color: #888;"># 1. Determine the best analyzer for this resource</span> analyzer = <span style="color: cyan;">self</span>.analyzers.get_for_resource(resource) <span style="color: #888;"># 2. Produce UKO triples using the most specific vocabulary</span> triples = analyzer.analyze(resource) <span style="color: #888;"># 3. Add provenance</span> <span style="color: magenta; font-weight: 600;">for</span> triple <span style="color: magenta; font-weight: 600;">in</span> triples: <span style="color: cyan;">self</span>._add_provenance(triple, resource) <span style="color: #888;"># 4. Store in graph backend</span> <span style="color: cyan;">self</span>.graph_backend.add_triples(triples) <span style="color: #888;"># 5. Also index into text and vector backends</span> <span style="color: cyan;">self</span>.text_indexer.index(resource, triples) <span style="color: cyan;">self</span>.vector_indexer.index(resource, triples)
Index Lifecycle
| Stage | Trigger | Action | Result |
|---|---|---|---|
| Resource added | agents resource add / agents project link-resource |
Immediate full indexing | All indices ready for instant search |
| Code changed | File modification detected | Immediate incremental update | Indices reflect latest state |
| Resource removed | agents resource remove / agents project unlink-resource |
Immediate cleanup | Stale data removed, UKO nodes marked historical |
| Maintenance | Scheduled | Full reindex, consistency check | Indices verified and optimized |
Fallback to Traditional Search
When advanced features are unavailable, the system gracefully degrades. The pipeline's ConfidenceWeightedSelector (default StrategySelector component) automatically routes to simpler strategies:
- Try
arce(requires all backends) -> if unavailable: - Try
breadth-depth-navigator(requires graph) -> if unavailable: - Try
semantic-embedding(requires vector) -> if unavailable: - Fall back to
simple-keyword(requires only text search / ripgrep)
ACMS Performance Characteristics
Assembly Latency
| Operation | Target Latency | Notes |
|---|---|---|
| Resource scope resolution | < 10ms | In-memory set operations |
| Scoped backend view creation | < 5ms | Filter injection |
Strategy can_handle check (all) |
< 20ms | Simple capability checks |
simple-keyword assembly |
< 100ms | Text search |
semantic-embedding assembly |
< 200ms | Vector search |
breadth-depth-navigator assembly |
< 500ms | Graph traversal + materialization |
arce assembly |
< 1s | Multi-modal search + ranking |
temporal-archaeology assembly |
< 2s | Cold tier queries |
| Fragment fusion pipeline (phases 2-3) | < 100ms | Dedup + depth resolution + scoring + knapsack + ordering + preamble |
| Total initial assembly | < 2s | Parallel strategy execution |
| Incremental refresh | < 500ms | Re-compression only |
Indexing and Query Performance
| Metric | Performance |
|---|---|
| Text indexing speed | 10,000 files/minute |
| Vector indexing speed | 1,000 files/minute (with GPU) |
| Graph indexing speed | 5,000 files/minute |
| Text search query | < 100ms for 1M files |
| Vector search query | < 200ms for 10M embeddings |
| Graph traversal query | < 500ms for 3-hop queries |
Scalability Guarantees
- 100K+ file projects: The scoping mechanism ensures strategies never see the full dataset. Scoped views filter at the backend level.
- Deep plan hierarchies (10+ levels): Skeleton compression ensures parent context overhead stays bounded (10% per level, multiplicatively compressed).
- Concurrent plans: Each plan gets independent ScopedBackendViews. No cross-plan interference.
- Dynamic budgets: The knapsack algorithm is O(n) in the number of fragments. Re-compression is O(1) per fragment.
- Maximum scale tested: 10M files, 1B+ triples, 100M+ embeddings.
Progressive Enhancement Path
Organizations can adopt ACMS features progressively. At each stage, existing resources are reindexed to take advantage of new capabilities:
| Stage | Features | Requirements | Benefit |
|---|---|---|---|
| Basic | Text search, file watching | ripgrep, sqlite | Instant exact-match search |
| Semantic | Vector embeddings, similarity search | Embedding model, vector DB | Instant semantic similarity search |
| Structural | UKO graph, relationship queries | Graph database, language analyzers | Instant relationship queries |
| Intelligent | ML-driven ranking, automated analysis | GPU, training data | Instant intelligent suggestions |
| Custom | Domain-specific ontologies, custom analyzers | Domain expertise | Instant domain-aware intelligence |
Summary of Timing:
- Indexing: Eager (happens immediately when resources are added/changed)
- Searching: Instant (because indices are pre-computed and ready)
- Sandboxing: Lazy (only when execution needs to modify a resource)
- Context Assembly: Real-time (but fast because it queries ready indices)
This design ensures agents never wait for index building during execution, providing a responsive and predictable experience even on massive codebases.
Storage and Persistence
!!! adr "Architecture Decision" The hybrid storage model, database strategy, and migration approach are defined in ADR-019: Storage and Persistence.
This section defines the complete storage strategy for every persistent concept in CleverAgents, specifying the storage backend, data format, lifecycle management, and migration strategy for each entity type.
Storage Architecture Overview
CleverAgents uses a hybrid storage model combining a relational database for structured data, the filesystem for configuration files and artifacts, and specialized indexes for search operations. The storage layer is accessed exclusively through repository interfaces defined in the domain layer, ensuring that storage backends can be swapped without affecting application logic.
@startuml
skinparam componentStyle rectangle
skinparam defaultFontSize 12
skinparam packageFontSize 13
skinparam packageFontStyle bold
package "Storage Architecture" as SA {
package "Relational Database" as RDB #E3F2FD {
component [Plans\nDecisions\nProjects\nResources\nSessions\nActors (db)\nInvariants\nChanges\nCheckpoint Metadata\nCorrection Attempts\nAudit Log] as Tables
}
package "Filesystem" as FS #C8E6C9 {
component [Config YAML Files] as Cfg
component [Sandbox Worktree Files] as Sbx
component [Artifact Storage] as Art
}
package "Search Indexes" as IDX #F3E5F5 {
component [Text (Tantivy)] as TI
component [Vector (FAISS)] as VI
component [Graph (Neo4j)] as GI
component [Checkpoint Store\n(filesystem)] as CS
}
}
@enduml
Entity Storage Map
The following table specifies where each CleverAgents concept is persisted, in both local and server modes:
| Entity | Local Storage | Server Storage | Format | Identifier |
|---|---|---|---|---|
| Plans | SQLite plans table |
Server PostgreSQL | Relational rows | ULID (plan_id) |
| Decisions | SQLite decisions table |
Server PostgreSQL | Relational rows + JSON context_snapshot |
ULID (decision_id) |
| Projects | SQLite projects table |
Server PostgreSQL (for non-local/ namespaces) |
Relational rows | Namespaced name |
| Resources | SQLite resources table + resource_links table |
Server PostgreSQL | Relational rows + DAG edges | ULID (resource_id) |
| Resource Types | SQLite resource_types table + YAML config files |
Server PostgreSQL + synced YAML | Relational rows | Namespaced name (built-ins unnamespaced) |
| Actions | SQLite actions table + YAML config files |
Server PostgreSQL + synced YAML | Relational rows | Namespaced name |
| Actors | SQLite actors table + YAML config files |
Server PostgreSQL + synced YAML | Relational rows + file reference | Namespaced name |
| Skills | SQLite skills table + YAML config files |
Server PostgreSQL + synced YAML | Relational rows + file reference | Namespaced name |
| Tools | SQLite tools table + YAML config files |
Server PostgreSQL + synced YAML | Relational rows + file reference | Namespaced name |
| Sessions | SQLite sessions table + session_messages table |
Server PostgreSQL | Relational rows | ULID (session_id) |
| Invariants | SQLite invariants table |
Server PostgreSQL | Relational rows | Auto-generated ID |
| Automation Profiles | SQLite automation_profiles table + YAML config files |
Server PostgreSQL + synced YAML | Relational rows | Namespaced name (built-ins unnamespaced) |
| Changes / ChangeSets | SQLite changes table |
Server PostgreSQL | Relational rows | ULID per change |
| Checkpoints | Filesystem (<data-dir>/checkpoints/<plan_id>/) + SQLite checkpoint_metadata table |
Server object storage + PostgreSQL metadata | Git refs or filesystem snapshots + metadata rows | ULID (checkpoint_id) |
| Correction Attempts | SQLite correction_attempts table + archived artifacts on filesystem |
Server PostgreSQL + object storage | Relational rows + file archives | ULID (correction_attempt_id) |
| Artifacts | Filesystem (<data-dir>/artifacts/<plan_id>/) |
Server object storage (S3-compatible) | Raw files (code, documents, images, etc.) | Filesystem path relative to plan |
| Context Indexes (Text) | Filesystem (<data-dir>/index/text/) |
Server-managed index cluster | Tantivy index files or SQLite FTS5 | Index directory per project |
| Context Indexes (Vector) | Filesystem (<data-dir>/index/vector/) |
Server-managed FAISS/Qdrant cluster | FAISS index files or Qdrant collections | Index directory per project |
| Context Indexes (Graph) | Neo4j database or rdflib files | Server-managed Neo4j instance | Graph triples | Graph database per project |
| Configuration | TOML file (<data-dir>/config.toml) |
N/A (server has own config) | TOML | Single file |
| Logs | Filesystem (<data-dir>/logs/) |
Server-managed log aggregation | Structured JSON (structlog) | Rotated log files |
| Audit Logs | SQLite audit_log table |
Server PostgreSQL | Relational rows | Auto-incrementing ID |
Database Schema Design
The relational database follows a normalized design with foreign key constraints enforcing referential integrity. The schema is version-controlled through Alembic migrations.
Key design decisions:
-
ULID primary keys: All time-series entities (plans, decisions, checkpoints, correction attempts) use ULID primary keys. This provides natural time-ordering without separate timestamp indexes, collision-free generation without coordination, and URL-safe string representation.
-
JSON columns for semi-structured data: Fields like
context_snapshot,alternatives_considered,artifacts_produced, andresource_bindingsare stored as JSON text columns. This avoids over-normalization for data that is always read and written as a unit, while still being queryable via SQLite'sjson_extract()or PostgreSQL'sjsonboperators. -
Soft delete pattern: Entities that support archival (actions, actors, skills) use a
statecolumn with values likeavailable,archived. No rows are physically deleted except during explicitagents initreset or retention policy enforcement. -
Optimistic concurrency control: The
updated_attimestamp column on mutable entities serves as a version check. Update operations includeWHERE updated_at = <previous_value>to detect concurrent modifications. On conflict, the operation fails with an explicit error rather than silently overwriting.
Core tables (SQLite DDL):
-- Plans table: the fundamental unit of orchestration CREATE TABLE plans ( plan_id TEXT PRIMARY KEY, -- ULID parent_plan_id TEXT REFERENCES plans(plan_id), root_plan_id TEXT NOT NULL REFERENCES plans(plan_id), action_name TEXT NOT NULL, -- namespaced action name phase TEXT NOT NULL DEFAULT 'strategize', -- action|strategize|execute|apply state TEXT NOT NULL DEFAULT 'queued', -- queued|processing|errored|complete|cancelled attempt INTEGER NOT NULL DEFAULT 1, automation_profile_name TEXT NOT NULL, effective_profile_snapshot TEXT NOT NULL, -- JSON: frozen profile at creation arguments TEXT, -- JSON: action arguments project_names TEXT NOT NULL, -- JSON array of project names strategy_actor_name TEXT, execution_actor_name TEXT, estimation_actor_name TEXT, invariant_actor_name TEXT, created_by TEXT, -- session ID or user identity created_at TEXT NOT NULL DEFAULT (strftime('%Y-%m-%dT%H:%M:%f', 'now')), updated_at TEXT NOT NULL DEFAULT (strftime('%Y-%m-%dT%H:%M:%f', 'now')), completed_at TEXT, error_message TEXT, error_traceback TEXT, cost_estimate_usd REAL, cost_actual_usd REAL, token_count_input INTEGER DEFAULT 0, token_count_output INTEGER DEFAULT 0 );CREATE INDEX idx_plans_parent ON plans(parent_plan_id); CREATE INDEX idx_plans_root ON plans(root_plan_id); CREATE INDEX idx_plans_phase_state ON plans(phase, state); CREATE INDEX idx_plans_action ON plans(action_name); CREATE INDEX idx_plans_created ON plans(created_at);
-- Decisions table: the decision tree CREATE TABLE decisions ( decision_id TEXT PRIMARY KEY, -- ULID plan_id TEXT NOT NULL REFERENCES plans(plan_id), parent_decision_id TEXT REFERENCES decisions(decision_id), decision_type TEXT NOT NULL, -- prompt_definition|invariant_enforced| -- strategy_choice|subplan_spawn| -- subplan_parallel_spawn|implementation_choice question TEXT NOT NULL, chosen_option TEXT NOT NULL, alternatives_considered TEXT, -- JSON array of alternative options confidence_score REAL, -- 0.0-1.0 rationale TEXT, context_snapshot TEXT NOT NULL, -- JSON: hot_context_hash, relevant_resources, -- actor_state_ref downstream_plan_ids TEXT, -- JSON array of child plan ULIDs artifacts_produced TEXT, -- JSON array of artifact references is_correction BOOLEAN DEFAULT FALSE, corrects_decision_id TEXT REFERENCES decisions(decision_id), superseded_by TEXT REFERENCES decisions(decision_id), sequence_number INTEGER NOT NULL, -- ordering within this plan created_at TEXT NOT NULL DEFAULT (strftime('%Y-%m-%dT%H:%M:%f', 'now')) );
CREATE INDEX idx_decisions_plan ON decisions(plan_id); CREATE INDEX idx_decisions_parent ON decisions(parent_decision_id); CREATE INDEX idx_decisions_type ON decisions(decision_type); CREATE INDEX idx_decisions_superseded ON decisions(superseded_by) WHERE superseded_by IS NOT NULL;
-- Resources table: the resource registry CREATE TABLE resources ( resource_id TEXT PRIMARY KEY, -- ULID name TEXT, -- namespaced name (NULL for auto-discovered children) resource_type_name TEXT NOT NULL, -- namespaced type name classification TEXT NOT NULL, -- physical|virtual description TEXT, properties TEXT, -- JSON: type-specific properties location TEXT, -- physical resources only content_hash TEXT, -- for equivalence tracking sandbox_strategy TEXT, -- override per resource created_at TEXT NOT NULL DEFAULT (strftime('%Y-%m-%dT%H:%M:%f', 'now')), updated_at TEXT NOT NULL DEFAULT (strftime('%Y-%m-%dT%H:%M:%f', 'now')) );
CREATE UNIQUE INDEX idx_resources_name ON resources(name) WHERE name IS NOT NULL; CREATE INDEX idx_resources_type ON resources(resource_type_name); CREATE INDEX idx_resources_classification ON resources(classification);
-- Resource DAG: parent/child relationships CREATE TABLE resource_links ( parent_id TEXT NOT NULL REFERENCES resources(resource_id), child_id TEXT NOT NULL REFERENCES resources(resource_id), link_type TEXT DEFAULT 'contains', -- contains|references|derived_from created_at TEXT NOT NULL DEFAULT (strftime('%Y-%m-%dT%H:%M:%f', 'now')), PRIMARY KEY (parent_id, child_id) );
CREATE INDEX idx_resource_links_child ON resource_links(child_id);
-- Project-resource linkage CREATE TABLE project_resources ( project_name TEXT NOT NULL, resource_id TEXT NOT NULL REFERENCES resources(resource_id), read_only BOOLEAN DEFAULT FALSE, linked_at TEXT NOT NULL DEFAULT (strftime('%Y-%m-%dT%H:%M:%f', 'now')), PRIMARY KEY (project_name, resource_id) );
-- Checkpoint metadata (actual data on filesystem) CREATE TABLE checkpoint_metadata ( checkpoint_id TEXT PRIMARY KEY, -- ULID plan_id TEXT NOT NULL REFERENCES plans(plan_id), decision_id TEXT REFERENCES decisions(decision_id), checkpoint_type TEXT NOT NULL, -- pre_write|post_step|manual resource_id TEXT REFERENCES resources(resource_id), filesystem_path TEXT NOT NULL, -- relative path within checkpoint dir size_bytes INTEGER, created_at TEXT NOT NULL DEFAULT (strftime('%Y-%m-%dT%H:%M:%f', 'now')) );
CREATE INDEX idx_checkpoints_plan ON checkpoint_metadata(plan_id);
-- Correction attempts CREATE TABLE correction_attempts ( correction_attempt_id TEXT PRIMARY KEY, -- ULID plan_id TEXT NOT NULL REFERENCES plans(plan_id), original_decision_id TEXT NOT NULL REFERENCES decisions(decision_id), new_decision_id TEXT REFERENCES decisions(decision_id), mode TEXT NOT NULL, -- revert|append guidance TEXT NOT NULL, archived_artifacts_path TEXT, -- filesystem path to archived originals state TEXT NOT NULL DEFAULT 'pending', -- pending|executing|complete|failed created_at TEXT NOT NULL DEFAULT (strftime('%Y-%m-%dT%H:%M:%f', 'now')), completed_at TEXT );
CREATE INDEX idx_corrections_plan ON correction_attempts(plan_id);
-- Audit log for apply operations CREATE TABLE audit_log ( id INTEGER PRIMARY KEY AUTOINCREMENT, event_type TEXT NOT NULL, -- plan_applied|plan_cancelled|resource_modified| -- correction_applied|config_changed plan_id TEXT, project_name TEXT, actor_name TEXT, user_identity TEXT, details TEXT NOT NULL, -- JSON: event-specific details created_at TEXT NOT NULL DEFAULT (strftime('%Y-%m-%dT%H:%M:%f', 'now')) );
CREATE INDEX idx_audit_event ON audit_log(event_type); CREATE INDEX idx_audit_plan ON audit_log(plan_id) WHERE plan_id IS NOT NULL; CREATE INDEX idx_audit_created ON audit_log(created_at);
Filesystem Layout
The data directory follows a deterministic layout. All paths are relative to core.data-dir (default: ~/.cleveragents):
~/.cleveragents/
├── config.toml # Global configuration
├── cleveragents.db # SQLite database (WAL mode)
├── cleveragents.db-wal # WAL file (auto-managed by SQLite)
├── cleveragents.db-shm # Shared memory file (auto-managed)
├── logs/
│ ├── cleveragents.log # Current log file (structured JSON)
│ ├── cleveragents.log.1 # Rotated log files
│ └── ...
├── cache/
│ ├── models/ # Cached model artifacts
│ ├── tools/ # Cached tool downloads
│ └── templates/ # Compiled Jinja2 templates
├── backups/
│ ├── 2026-02-08T12-30-00/ # Timestamped backup snapshots
│ └── ...
├── checkpoints/
│ ├── 01HXR1C1D2E3F4G5H6I7.../ # Per-plan checkpoint directories
│ │ ├── chk_01HXR1E1.tar.gz # Compressed checkpoint snapshots
│ │ └── ...
│ └── ...
├── artifacts/
│ ├── 01HXR1C1D2E3F4G5H6I7.../ # Per-plan artifact directories
│ │ ├── src/routes/health.py # Generated/modified artifacts
│ │ └── ...
│ └── ...
├── index/
│ ├── text/
│ │ ├── local_api-service/ # Per-project Tantivy indexes
│ │ └── ...
│ ├── vector/
│ │ ├── local_api-service/ # Per-project FAISS indexes
│ │ └── ...
│ └── graph/
│ └── ... # rdflib stores or Neo4j connection info
├── sessions/
│ └── ... # Session state files (if needed beyond DB)
└── contexts/
└── ... # Exported/cached context snapshots
Data Lifecycle and Retention
| Data Type | Retention Policy | Cleanup Trigger |
|---|---|---|
| Plans (active) | Indefinite until the plan reaches applied or cancelled state |
Manual deletion |
| Plans (terminal) | Governed by core.backup.retention-days for archives |
Automatic cleanup daemon |
| Decisions | Retained as long as parent plan exists | Cascade delete with plan |
| Checkpoints | sandbox.checkpoint.max-per-plan per plan; oldest pruned first (keeping first + most recent) |
After apply or on retention limit |
| Correction archives | core.backup.retention-days after plan reaches applied state |
Automatic cleanup daemon |
| Logs | core.log.retention-days |
Daily rotation + age-based cleanup |
| Backups | core.backup.retention-days |
Age-based cleanup |
| Cache | No automatic retention; manual cleanup via agents diagnostics --cleanup |
User-initiated |
| Search indexes | Rebuilt on resource add/remove; no explicit retention | Re-index on change |
| Audit logs | Indefinite (critical for compliance) | Manual archival only |
Migration Strategy
Database schema changes are managed through Alembic migrations with the following rules:
- Forward-only in production: Downgrade scripts are provided but only used in development. Production deployments only move forward.
- Auto-apply on startup: When
CLEVERAGENTS_AUTO_APPLY_MIGRATIONSis set (default in development), migrations run automatically on first database access. In production, migrations are triggered byagents initor explicit upgrade commands. - Backward compatibility window: Each migration maintains backward compatibility with the previous schema version for at least one minor release cycle, allowing rolling upgrades in server mode.
- Data migration separation: Schema DDL changes and data transformations are kept in separate migration steps to enable rollback of data changes independently.
Observability
!!! adr "Architecture Decision" The observability strategy, structured logging, metrics, and distributed tracing are defined in ADR-025: Observability and Logging.
CleverAgents implements a comprehensive observability strategy across three pillars — logging, metrics, and tracing — designed to provide full visibility into plan execution, actor behavior, tool invocations, and resource access. Every significant operation in the system is observable, enabling debugging of complex multi-plan hierarchies and performance optimization.
Structured Logging
All logging uses structlog with JSON output format. Every log entry includes contextual fields that enable correlation across plan hierarchies:
import structloglogger = structlog.get_logger()
# Every log entry automatically includes bound context logger = logger.bind( plan_id="01HXR1C1D2E3F4G5H6I7J8K9L0", phase="execute", actor_name="anthropic/claude-3.5-sonnet", session_id="01HXM2A6K1P2E9Q9D4GQ7J4S7Z", )
# Example: tool invocation logging logger.info( "tool_invoked", tool_name="local/write-file", skill_name="local/file-ops", resource_id="01HXR1A1B2C3D4E5F6G7H8J9K0", resource_type="git-checkout", input_schema_hash="sha256:a1b2c3...", duration_ms=142, writes=True, checkpoint_created=True, )
Log levels and their semantics:
| Level | Usage | Examples |
|---|---|---|
DEBUG |
Internal state transitions, context building details, LLM prompt construction | "Building hot context: 12 files, 8,420 tokens", "Decision confidence computed: 0.87" |
INFO |
Significant lifecycle events, phase transitions, tool completions | "Plan entered execute phase", "Tool write-file completed in 142ms", "Validation passed: pytest" |
WARNING |
Degraded operations, approaching limits, non-fatal issues | "Budget at 82% ($4.10/$5.00)", "Checkpoint pruned: exceeded max-per-plan", "Provider fallback: openai -> anthropic" |
ERROR |
Operation failures, invariant violations, unrecoverable tool errors | "Validation failed: pytest exited 1", "Invariant violated: API compatibility", "LLM provider error: rate limited" |
Log correlation: Every log entry within a plan execution context includes plan_id, root_plan_id, parent_plan_id, phase, state, and attempt. This enables tracing a single decision through the entire plan hierarchy using standard log querying tools (e.g., jq '.plan_id == "01HXR1C1..."').
Event System
CleverAgents emits domain events for every significant state change. Events flow through two channels:
- RxPY Reactive Stream: Real-time event stream for in-process subscribers (TUI updates, progress bars, streaming output). Uses
Subjectfor fan-out distribution. - Persistent Event Log: All events are written to the
audit_logtable for post-hoc analysis and compliance.
Event taxonomy:
from enum import Enum from pydantic import BaseModel from datetime import datetimeclass EventType(str, Enum): # Plan lifecycle events PLAN_CREATED = "plan.created" PLAN_PHASE_CHANGED = "plan.phase_changed" PLAN_STATE_CHANGED = "plan.state_changed" PLAN_APPLIED = "plan.applied" PLAN_CANCELLED = "plan.cancelled" PLAN_ERRORED = "plan.errored"
<span style="color: #888;"># Decision events</span> DECISION_CREATED = <span style="color: #66cc66;">"decision.created"</span> DECISION_APPROVED = <span style="color: #66cc66;">"decision.approved"</span> DECISION_CORRECTED = <span style="color: #66cc66;">"decision.corrected"</span> DECISION_SUPERSEDED = <span style="color: #66cc66;">"decision.superseded"</span> <span style="color: #888;"># Invariant events</span> INVARIANT_RECONCILED = <span style="color: #66cc66;">"invariant.reconciled"</span> INVARIANT_VIOLATED = <span style="color: #66cc66;">"invariant.violated"</span> INVARIANT_ENFORCED = <span style="color: #66cc66;">"invariant.enforced"</span> <span style="color: #888;"># Actor events</span> ACTOR_INVOKED = <span style="color: #66cc66;">"actor.invoked"</span> ACTOR_COMPLETED = <span style="color: #66cc66;">"actor.completed"</span> ACTOR_ERRORED = <span style="color: #66cc66;">"actor.errored"</span> ACTOR_ESCALATED = <span style="color: #66cc66;">"actor.escalated"</span> <span style="color: #888;"># confidence below threshold</span> <span style="color: #888;"># Tool events</span> TOOL_INVOKED = <span style="color: #66cc66;">"tool.invoked"</span> TOOL_COMPLETED = <span style="color: #66cc66;">"tool.completed"</span> TOOL_ERRORED = <span style="color: #66cc66;">"tool.errored"</span> TOOL_RETRIED = <span style="color: #66cc66;">"tool.retried"</span> <span style="color: #888;"># Resource events</span> RESOURCE_ACCESSED = <span style="color: #66cc66;">"resource.accessed"</span> RESOURCE_MODIFIED = <span style="color: #66cc66;">"resource.modified"</span> RESOURCE_INDEXED = <span style="color: #66cc66;">"resource.indexed"</span> <span style="color: #888;"># Sandbox events</span> SANDBOX_CREATED = <span style="color: #66cc66;">"sandbox.created"</span> SANDBOX_COMMITTED = <span style="color: #66cc66;">"sandbox.committed"</span> SANDBOX_ROLLED_BACK = <span style="color: #66cc66;">"sandbox.rolled_back"</span> CHECKPOINT_CREATED = <span style="color: #66cc66;">"checkpoint.created"</span> CHECKPOINT_RESTORED = <span style="color: #66cc66;">"checkpoint.restored"</span> <span style="color: #888;"># Context events</span> CONTEXT_BUILT = <span style="color: #66cc66;">"context.built"</span> CONTEXT_QUERY_EXECUTED = <span style="color: #66cc66;">"context.query_executed"</span> <span style="color: #888;"># Validation events</span> VALIDATION_STARTED = <span style="color: #66cc66;">"validation.started"</span> VALIDATION_PASSED = <span style="color: #66cc66;">"validation.passed"</span> VALIDATION_FAILED = <span style="color: #66cc66;">"validation.failed"</span> <span style="color: #888;"># Session events</span> SESSION_CREATED = <span style="color: #66cc66;">"session.created"</span> SESSION_MESSAGE_SENT = <span style="color: #66cc66;">"session.message_sent"</span> <span style="color: #888;"># Cost events</span> BUDGET_WARNING = <span style="color: #66cc66;">"budget.warning"</span> BUDGET_EXCEEDED = <span style="color: #66cc66;">"budget.exceeded"</span>
class DomainEvent(BaseModel): """Base event model emitted by all domain operations.""" event_type: EventType timestamp: datetime plan_id: str | None = None root_plan_id: str | None = None session_id: str | None = None actor_name: str | None = None project_name: str | None = None details: dict # Event-type-specific payload
Event emission pattern: Domain services emit events through an injected EventBus interface. The infrastructure layer provides two implementations:
from typing import Protocol import rx from rx.subject import Subjectclass EventBus(Protocol): def emit(self, event: DomainEvent) -> None: ... def subscribe(self, event_type: EventType, handler: Callable) -> None: ...
class ReactiveEventBus: """In-process event bus using RxPY for real-time distribution.""" def init(self): self._subject = Subject() self._subscriptions: dict[EventType, list[Callable]] = {}
<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">emit</span>(<span style="color: cyan;">self</span>, event: DomainEvent) -> <span style="color: cyan;">None</span>: <span style="color: #888;"># Push to reactive stream for real-time subscribers</span> <span style="color: cyan;">self</span>._subject.on_next(event) <span style="color: #888;"># Persist to audit log</span> <span style="color: cyan;">self</span>._persist_audit(event) <span style="color: #888;"># Dispatch to type-specific handlers</span> <span style="color: magenta; font-weight: 600;">for</span> handler <span style="color: magenta; font-weight: 600;">in</span> <span style="color: cyan;">self</span>._subscriptions.get(event.event_type, []): handler(event) <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">subscribe</span>(<span style="color: cyan;">self</span>, event_type: <span style="color: cyan;">EventType</span>, handler: <span style="color: cyan;">Callable</span>) -> <span style="color: cyan;">None</span>: <span style="color: cyan;">self</span>._subscriptions.setdefault(event_type, []).append(handler) <span style="color: cyan;">@property</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">stream</span>(<span style="color: cyan;">self</span>) -> rx.Observable: <span style="color: #888;">"""Raw observable stream for advanced RxPY operators."""</span> <span style="color: magenta; font-weight: 600;">return</span> <span style="color: cyan;">self</span>._subject
LLM Call Tracing
Every LLM invocation is traced with full context for debugging and cost tracking:
# Trace record for each LLM call
class LLMTrace(BaseModel):
trace_id: str # ULID
plan_id: str
decision_id: str | None
actor_name: str
provider: str # openai, anthropic, google, etc.
model: str # gpt-4o, claude-3.5-sonnet, etc.
prompt_tokens: int
completion_tokens: int
total_tokens: int
cost_usd: float
latency_ms: int
temperature: float
tool_calls: list[str] # tool names invoked by the LLM
context_hash: str # hash of the context window contents
context_refs: list[str] # resource IDs referenced in context
streaming: bool
retry_count: int
error: str | None
timestamp: datetime
LangSmith integration: When CLEVERAGENTS_LANGSMITH_ENABLED is true, all LLM traces are additionally forwarded to LangSmith for visualization, comparison, and cost analysis. The integration maps CleverAgents plan IDs to LangSmith run tags, enabling filtering by plan hierarchy in the LangSmith UI.
Metrics Collection
CleverAgents collects operational metrics for performance monitoring and capacity planning. Metrics are emitted as structured log entries (for local mode) and optionally exported to Prometheus (for server mode):
| Metric | Type | Description |
|---|---|---|
plan.duration_seconds |
Histogram | Total wall-clock time per plan, labeled by phase |
plan.cost_usd |
Counter | Cumulative API cost per plan |
plan.decisions_count |
Counter | Number of decisions per plan |
plan.child_plans_count |
Counter | Number of child plans spawned |
actor.invocation_duration_ms |
Histogram | Per-actor invocation latency |
actor.token_usage |
Counter | Token counts by provider and model |
tool.invocation_duration_ms |
Histogram | Per-tool invocation latency |
tool.error_rate |
Counter | Tool invocation failures by tool name |
context.build_duration_ms |
Histogram | Context building time by tier (hot/warm/cold) |
context.tokens_used |
Gauge | Current token usage in hot context |
index.query_duration_ms |
Histogram | Index query latency by backend (text/vector/graph) |
sandbox.operation_duration_ms |
Histogram | Sandbox create/commit/rollback time |
validation.duration_seconds |
Histogram | Per-validation tool invocation execution time |
validation.pass_rate |
Counter | Validation pass/fail counts (from passed field in return JSON) |
Diagnostic Dashboard
The agents diagnostics command provides a real-time health check combining all observability data:
- System health: Config file readability, database accessibility, disk space, provider API key validation.
- Index health: Text index status, vector index dimensionality and count, graph store connectivity.
- Active plan overview: Running plans, queued plans, resource utilization.
- Cost summary: Current session cost, budget utilization, provider-level breakdown.
- Performance summary: Average plan duration, tool call latency percentiles, context build times.
Security Model
!!! adr "Architecture Decision" Sandbox isolation is covered in ADR-015: Sandbox and Checkpoint. The server-mode security model is defined in ADR-023: Server Mode.
CleverAgents implements a defense-in-depth security model addressing five concerns: sandbox isolation, access control, prompt injection mitigation, secret management, and audit logging. The security model differs between local mode (single-user, trusted environment) and server mode (multi-user, potentially untrusted environment).
Sandbox Isolation
The sandbox is the primary safety mechanism preventing untested changes from reaching production resources. Every plan's Execute phase runs within an isolated sandbox unless explicitly disabled by the automation profile (require_sandbox: false).
Sandbox strategies and their security properties:
| Strategy | Isolation Level | Resource Type | Mechanism | Rollback |
|---|---|---|---|---|
git_worktree |
Process-level filesystem isolation | Git repositories | Creates a separate git worktree on a plan-specific branch. Changes are confined to the worktree directory. Merge back to main requires explicit apply. | git checkout -- . or worktree deletion |
filesystem_copy |
Process-level filesystem isolation | Non-git filesystems | Creates a copy-on-write snapshot (using cp --reflink=auto where supported, falling back to full copy). Original files are never modified during execution. |
Delete the copy directory |
transaction_rollback |
Database transaction isolation | Databases | Opens a database transaction with SERIALIZABLE isolation level. All changes are staged within the transaction. Commit on apply; rollback on cancel/error. |
ROLLBACK |
none |
No isolation | Any | Changes are applied directly. Requires require_sandbox: false in the automation profile. Only appropriate when external safety mechanisms exist. |
No automatic rollback |
Sandbox security invariants:
- No cross-sandbox access: Each plan's sandbox is isolated. Plan A cannot read or write Plan B's sandbox contents. The sandbox path is derived deterministically from the plan ULID, preventing path traversal.
- Read-only source guarantee during Execute: The original project resources are never modified during Execute. All tool write operations target the sandbox copy.
- Atomic apply: The Apply phase is an atomic operation — either all sandbox changes are committed to the real resources, or none are. For git-based resources, this uses git merge. For database resources, this uses transaction commit.
- Sandbox cleanup: Sandbox directories are cleaned up according to
sandbox.cleanuppolicy. Sensitive data (API responses, LLM outputs) is purged from the sandbox on cleanup.
Prompt Injection Mitigation
In server mode, where user-provided content may flow into LLM prompts, CleverAgents implements several mitigations:
-
Input sanitization: User-provided text in action arguments, invariant text, and session prompts is sanitized before inclusion in LLM prompts. HTML entities, control characters, and known injection patterns are escaped or rejected.
-
Prompt boundary markers: System prompts and user content are separated by clear boundary markers that the LLM is instructed to recognize. The system prompt explicitly states that content between
[USER_CONTENT_START]and[USER_CONTENT_END]markers is user-provided and should not be treated as system instructions. -
Output validation: LLM outputs that are used as tool invocations are validated against the tool's JSON Schema before execution. This prevents the LLM from being tricked into invoking tools with malicious parameters.
-
Tool capability restrictions: Tools declare their capabilities (
read_only,writes,checkpointable,side_effects). The execution engine enforces these declarations — a tool declared asread_onlycannot invoke write operations even if the LLM requests it. -
Unsafe tool gating: Tools marked as
unsafe(e.g., arbitrary shell execution, network access) are blocked unless the automation profile explicitly setsallow_unsafe_tools: true. In server mode, only administrators can create automation profiles that allow unsafe tools.
Secret Management
Secrets (API keys, database credentials, authentication tokens) are managed through a layered approach:
-
Environment variables (preferred): Provider API keys (
OPENAI_API_KEY,ANTHROPIC_API_KEY, etc.) and the server token (CLEVERAGENTS_SERVER_TOKEN) are read from environment variables. This is the recommended approach for both local and CI environments. -
Configuration file (fallback): API keys can be stored in
config.tomlunder theprovider.*section. The config file should have restrictive permissions (chmod 600). Theagents diagnosticscommand warns if config file permissions are too permissive. -
Secret masking in logs: All log output and CLI display automatically masks values that match known secret patterns (keys starting with
sk-,sk-ant-,tok_, etc.). Masked values are replaced with***REDACTED***. The masking is applied at the structlog processor level, ensuring no secret leaks through any log path. -
Secret masking in LLM context: Before constructing LLM prompts, the context builder scans for patterns matching known secret formats and replaces them with
[REDACTED]. This prevents accidental exposure of secrets in LLM training data (for providers that use customer data for training, which CleverAgents discourages). -
Server mode credential isolation: In server mode, each user's API keys are stored encrypted in the server database using AES-256-GCM with a per-user key derived from the server's master secret. Keys are decrypted only at the moment of LLM invocation and are never transmitted to the client CLI.
Audit Logging
All security-relevant operations are recorded in the audit_log table with the following events:
| Event Type | Trigger | Details Captured |
|---|---|---|
plan_applied |
agents plan apply |
Plan ID, project names, files changed, validation results, user identity |
plan_cancelled |
agents plan cancel |
Plan ID, reason, resources released |
resource_modified |
Tool write operations during Execute | Resource ID, modification type, sandbox path, tool name |
correction_applied |
agents plan correct |
Correction attempt ID, original decision ID, mode, guidance |
config_changed |
agents config set |
Key, old value (masked if secret), new value (masked if secret), user identity |
entity_deleted |
agents <entity> delete/remove |
Entity type, entity name, cascade effects |
session_created |
agents session create |
Session ID, actor name, user identity |
auth_success |
Server mode login | User identity, IP address, token prefix |
auth_failure |
Server mode failed login | Attempted identity, IP address, failure reason |
Audit log retention: Audit logs are never automatically deleted. In server mode, audit logs can be exported for compliance archival using server administration commands. In local mode, audit logs are preserved in the SQLite database indefinitely and are included in backup snapshots.
Extensibility
!!! adr "Architecture Decision" The extensibility model and plugin architecture are grounded in the layered design defined in ADR-001: Layered Architecture and the configuration system in ADR-024: Configuration System.
CleverAgents is designed as an extensible platform where every major subsystem supports custom implementations. The extensibility model follows the Open/Closed Principle — the system is open for extension through well-defined interfaces, but closed for modification of core behavior.
Plugin Architecture Overview
@startuml
skinparam componentStyle rectangle
skinparam defaultFontSize 12
skinparam packageFontSize 13
skinparam packageFontStyle bold
skinparam componentFontSize 11
package "Extension Points" as EP {
package "User-Facing Extensions" as UFE #C8E6C9 {
component [Custom Tools\n(YAML/Python)] as CT
component [Custom Skills\n(YAML)] as CS
component [Custom Actors\n(YAML/Graph)] as CA
}
package "Integration Extensions" as IE #E3F2FD {
component [MCP Servers\n(stdio/SSE)] as MCP
component [Agent Skills\nStandard] as ASS
component [Custom LLM\nProviders] as CLP
}
package "Core Registries" as CR #F3E5F5 {
component [Tool Registry\nSkill Registry\nActor Registry\nProvider Registry] as REG
}
package "Infrastructure Extensions" as INF #FFF9C4 {
component [Custom Resource\nTypes (YAML)] as CRT
component [Custom Sandbox\nStrategies (Python)] as CSS
component [Custom Index\nBackends] as CIB
}
package "ACMS Pipeline Extensions" as ACMS #E8EAF6 {
component [Custom UKO\nAnalyzers / Vocabularies] as UKO
component [Custom Context\nStrategies] as CCS
component [Custom Pipeline\nComponents (10 slots)] as CPC
}
CT -down-> REG
CS -down-> REG
CA -down-> REG
MCP -down-> REG
ASS -down-> REG
CLP -down-> REG
REG -down-> CRT
REG -down-> CSS
REG -down-> CIB
REG -down-> UKO
REG -down-> CCS
REG -down-> CPC
}
@enduml
Tool Extensibility
Tools are the atomic unit of execution and the primary extension point. There are five ways to add tools:
-
Custom Python tools: Define a tool with inline Python code in a YAML configuration file. The code runs in a sandboxed execution context with access to the plan context, resource bindings, and checkpoint API.
# File: tools/my-custom-tool.yaml tool: name: local/analyze-complexity description: "Compute cyclomatic complexity for Python files" source: custominput_schema: type: object properties: file_path: { type: string } required: [file_path]
output_schema: type: object properties: complexity: { type: integer } functions: { type: array }
capability: writes: false checkpointable: false
resource_slots: - name: source type: git-checkout binding: contextual
code: | import ast content = ctx.resources["source"].read(params["file_path"]) tree = ast.parse(content) # ... complexity analysis ... return {"complexity": total, "functions": results} -
MCP server tools: Any MCP-compliant server can expose tools to CleverAgents. Skills reference MCP servers by their transport configuration:
# File: skills/kubernetes-ops.yaml skill: name: local/kubernetes-ops description: "Kubernetes cluster management tools" mcp_servers: - transport: stdio command: npx args: ["-y", "@anthropic/mcp-kubernetes"] -
Agent Skills Standard tools: Tools organized in standard folder structures are auto-discovered and registered:
skill: name: local/project-tools agent_skills_dirs: - ./agent-skills/ -
Built-in tools: Core file operations (
read_file,write_file,edit_file,delete_file,move_file,list_files,search_files), plan operations (create-subplan), and system operations are provided as built-in tools grouped into built-in skills. -
Anonymous inline tools: Tools can be defined inline within actor graph nodes or skill definitions for one-off operations that don't need global registration.
Tool lifecycle hooks: Every tool follows a four-phase lifecycle — discover (find available capabilities), activate (prepare for use, acquire connections), execute (perform the operation), deactivate (release resources). The tool registry manages these lifecycle transitions.
Skill Extensibility
Skills compose tools into capability collections. Skills support:
- Tool aggregation: Reference named tools from the Tool Registry
- Inline anonymous tools: Define tools directly in the skill YAML
- Skill inclusion: Include other skills by reference (composability)
- MCP server aggregation: Expose tools from one or more MCP servers
- Agent Skills Standard: Expose tools from Agent Skills Standard directories
Actor Extensibility
Actors can be extended through:
- YAML-defined agents: Single LLM actors with custom system prompts, temperature settings, tool bindings, and capability constraints.
- YAML-defined graphs: LangGraph topologies with multiple actors and tool nodes connected by edges, conditional routing, and parallel execution groups.
- Provider extension: New LLM providers can be added by implementing the
AIProviderInterfaceprotocol and registering with theProviderRegistry. The registry uses auto-discovery to detect installedlangchain-*packages.
from typing import Protocolclass AIProviderInterface(Protocol): """Protocol for LLM provider implementations."""
<span style="color: cyan;">@property</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">provider_name</span>(<span style="color: cyan;">self</span>) -> <span style="color: cyan;">str</span>: ... <span style="color: cyan;">@property</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">capabilities</span>(<span style="color: cyan;">self</span>) -> ProviderCapabilities: ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">create_chat_model</span>( <span style="color: cyan;">self</span>, model: <span style="color: cyan;">str</span>, temperature: <span style="color: cyan;">float</span> = <span style="color: yellow;">0.7</span>, **kwargs, ) -> BaseChatModel: ... <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">create_embedding_model</span>( <span style="color: cyan;">self</span>, model: <span style="color: cyan;">str</span>, **kwargs, ) -> BaseEmbeddings: ...
Custom Resource Types
New resource types extend CleverAgents to manage any kind of resource:
# File: resource-types/docker-registry.yaml resource_type: name: local/docker-registry description: "Docker container registry" classification: physical user_addable: truecli_args: - name: registry_url type: string required: true - name: repository type: string required: true - name: tag type: string default: "latest"
sandbox_strategy: none # Docker images are immutable handler: docker_registry_handler
child_types: - type: local/docker-image relationship: contains auto_discover: true
auto_discovery: enabled: true strategy: list_tags
Custom Index Backends
The indexing layer is abstracted behind three interfaces (text, vector, graph). Custom implementations can be registered for each:
class TextIndexBackend(Protocol): """Interface for full-text search backends.""" def index_document(self, project: str, doc_id: str, content: str, metadata: dict) -> None: ... def search(self, project: str, query: str, limit: int) -> list[SearchResult]: ... def remove_document(self, project: str, doc_id: str) -> None: ... def rebuild_index(self, project: str) -> None: ...class VectorIndexBackend(Protocol): """Interface for vector similarity search backends.""" def index_embedding(self, project: str, doc_id: str, embedding: list[float], metadata: dict) -> None: ... def search_similar(self, project: str, query_embedding: list[float], limit: int, min_relevance: float) -> list[VectorResult]: ... def remove_embedding(self, project: str, doc_id: str) -> None: ...
class GraphIndexBackend(Protocol): """Interface for knowledge graph backends.""" def add_triple(self, project: str, subject: str, predicate: str, obj: str) -> None: ... def query(self, project: str, sparql: str) -> list[dict]: ... def remove_triples(self, project: str, subject: str | None, predicate: str | None, obj: str | None) -> None: ...
New backends are registered via configuration:
# config.toml [index.text] backend = "custom" custom_module = "my_extensions.elasticsearch_backend" custom_class = "ElasticsearchTextIndex"
[index.text.custom_options] hosts = ["http://localhost:9200"] index_prefix = "cleveragents"
Custom Sandbox Strategies
The sandbox layer supports custom isolation strategies for specialized resource types:
class SandboxStrategy(Protocol):
"""Interface for sandbox isolation strategies."""
def create(self, plan_id: str, resource: Resource) -> SandboxRef: ...
def read(self, ref: SandboxRef, path: str) -> bytes: ...
def write(self, ref: SandboxRef, path: str, content: bytes) -> Change: ...
def diff(self, ref: SandboxRef) -> DiffView: ...
def commit(self, ref: SandboxRef) -> None: ...
def rollback(self, ref: SandboxRef) -> None: ...
def checkpoint(self, ref: SandboxRef, checkpoint_id: str) -> None: ...
def restore_checkpoint(self, ref: SandboxRef, checkpoint_id: str) -> None: ...
def cleanup(self, ref: SandboxRef) -> None: ...
Custom strategies are mapped to resource types via the resource type configuration's sandbox_strategy field, or globally via sandbox.strategy config.
ACMS Extensions
The ACMS provides five categories of extension points: UKO analyzers, index backends, context strategies, UKO vocabularies, and Context Assembly Pipeline components.
Domain and Language-Specific Analyzers
Analyzers are pluggable components that parse resources and produce UKO triples. They are organized by domain (code, documents, data, infrastructure) and then by specific technology within each domain:
PluginSystem: analyzers: # --- Source Code Analyzers (Layer 1: uko-code + Layer 2/3) --- python: class: "PythonAnalyzer" features: [AST parsing, Type inference via mypy, Import resolution, Docstring extraction] typescript: class: "TypeScriptAnalyzer" features: [TSC-based parsing, Type extraction, Module resolution, JSDoc parsing] rust: class: "RustAnalyzer" features: [rust-analyzer integration, Lifetime analysis, Trait resolution, Macro expansion] custom_dsl: class: "CustomDSLAnalyzer" config: grammar: "path/to/grammar.peg" semantic_rules: "path/to/rules.yaml"<span style="color: #888;"># --- Document Analyzers (Layer 1: uko-doc) ---</span> <span style="color: cyan; font-weight: 600;">markdown</span>: <span style="color: cyan; font-weight: 600;">class</span>: <span style="color: #66cc66;">"MarkdownAnalyzer"</span> <span style="color: cyan; font-weight: 600;">features</span>: [<span style="color: #66cc66;">Heading hierarchy, Link extraction, Code block detection, Topic inference, Cross-reference resolution</span>] <span style="color: cyan; font-weight: 600;">restructuredtext</span>: <span style="color: cyan; font-weight: 600;">class</span>: <span style="color: #66cc66;">"ReStructuredTextAnalyzer"</span> <span style="color: cyan; font-weight: 600;">features</span>: [<span style="color: #66cc66;">Directive parsing, Role resolution, Cross-reference resolution, Index extraction</span>] <span style="color: cyan; font-weight: 600;">html</span>: <span style="color: cyan; font-weight: 600;">class</span>: <span style="color: #66cc66;">"HTMLDocumentAnalyzer"</span> <span style="color: cyan; font-weight: 600;">features</span>: [<span style="color: #66cc66;">Semantic HTML parsing, Heading extraction, Link graph, Readability scoring</span>] <span style="color: #888;"># --- Data Schema Analyzers (Layer 1: uko-data) ---</span> <span style="color: cyan; font-weight: 600;">postgresql</span>: <span style="color: cyan; font-weight: 600;">class</span>: <span style="color: #66cc66;">"PostgreSQLAnalyzer"</span> <span style="color: cyan; font-weight: 600;">features</span>: [<span style="color: #66cc66;">Schema introspection, DDL extraction, Foreign key graph, View dependency analysis, Stored procedure parsing, Statistics collection</span>] <span style="color: cyan; font-weight: 600;">mysql</span>: <span style="color: cyan; font-weight: 600;">class</span>: <span style="color: #66cc66;">"MySQLAnalyzer"</span> <span style="color: cyan; font-weight: 600;">features</span>: [<span style="color: #66cc66;">Schema introspection, DDL extraction, Foreign key graph, Trigger parsing</span>] <span style="color: cyan; font-weight: 600;">sqlite</span>: <span style="color: cyan; font-weight: 600;">class</span>: <span style="color: #66cc66;">"SQLiteAnalyzer"</span> <span style="color: cyan; font-weight: 600;">features</span>: [<span style="color: #66cc66;">Schema introspection, DDL extraction, Virtual table detection</span>] <span style="color: #888;"># --- Infrastructure Analyzers (Layer 1: uko-infra) ---</span> <span style="color: cyan; font-weight: 600;">docker_compose</span>: <span style="color: cyan; font-weight: 600;">class</span>: <span style="color: #66cc66;">"DockerComposeAnalyzer"</span> <span style="color: cyan; font-weight: 600;">features</span>: [<span style="color: #66cc66;">Service graph, Port mapping, Volume resolution, Environment variable extraction</span>] <span style="color: cyan; font-weight: 600;">kubernetes</span>: <span style="color: cyan; font-weight: 600;">class</span>: <span style="color: #66cc66;">"KubernetesAnalyzer"</span> <span style="color: cyan; font-weight: 600;">features</span>: [<span style="color: #66cc66;">Resource parsing, Service dependency graph, ConfigMap/Secret references, Ingress routing</span>]
Index Backend Providers
backends:
graph:
- { name: "blazegraph", class: "BlazegraphBackend", features: ["SPARQL", "reasoning"] }
- { name: "neo4j", class: "Neo4jBackend", features: ["Cypher", "APOC", "GDS"] }
vector:
- { name: "faiss", class: "FaissBackend", features: ["GPU acceleration", "HNSW"] }
- { name: "qdrant", class: "QdrantBackend", features: ["filtering", "payloads"] }
embedders:
- { name: "openai", class: "OpenAIEmbedder", models: ["text-embedding-3-large"] }
- { name: "local", class: "LocalEmbedder", models: ["all-MiniLM-L6-v2"] }
Adding a New UKO Vocabulary (e.g., for Terraform)
@prefix uko-tf: <https://cleveragents.ai/ontology/uko/infra/terraform#> .
uko-tf:TerraformModule a owl:Class ; rdfs:subClassOf uko-infra:ConfigBlock, uko:Container . uko-tf:Resource a owl:Class ; rdfs:subClassOf uko:Container . uko-tf:Variable a owl:Class ; rdfs:subClassOf uko-infra:ConfigKey . uko-tf:Output a owl:Class ; rdfs:subClassOf uko:Boundary . uko-tf:provisionsOn a owl:ObjectProperty ; rdfs:subPropertyOf uko:dependsOn ; rdfs:domain uko-tf:Resource ; rdfs:range uko-infra:Service .
Adding a New Context Strategy
class MyDomainStrategy: name = "my-namespace/domain-strategy" capabilities = StrategyCapabilities( uses_graph=True, uses_text=True, uko_levels=["uko:", "uko-tf:"], resource_types=["*"], supports_depth_breadth=True, quality_score=0.8, )<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">can_handle</span>(<span style="color: cyan;">self</span>, request, backends): <span style="color: magenta; font-weight: 600;">if</span> any(<span style="color: #66cc66;">"terraform"</span> <span style="color: magenta; font-weight: 600;">in</span> f <span style="color: magenta; font-weight: 600;">for</span> f <span style="color: magenta; font-weight: 600;">in</span> request.focus): <span style="color: magenta; font-weight: 600;">return</span> <span style="color: yellow;">0.9</span> <span style="color: magenta; font-weight: 600;">return</span> <span style="color: yellow;">0.0</span> <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">assemble</span>(<span style="color: cyan;">self</span>, request, backends, budget, plan_context): <span style="color: #888;"># Custom logic...</span> ...
Register via config:
[context.strategies.custom]
"my-namespace/domain-strategy" = "my_package.strategies.MyDomainStrategy"
Replacing Pipeline Components
Any of the 10 Context Assembly Pipeline components can be replaced by implementing the corresponding Protocol and registering the implementation. Components can be overridden at three scopes (plan > project > global), with more specific scopes taking precedence.
class MyCustomScorer: """A domain-aware fragment scorer that boosts infrastructure fragments when the plan focuses on deployment tasks."""<span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">score</span>(<span style="color: cyan;">self</span>, fragments: <span style="color: cyan;">list</span>[ContextFragment], plan_context: PlanContext) -> <span style="color: cyan;">list</span>[ScoredFragment]: scored = [] is_deploy = <span style="color: #66cc66;">"deploy"</span> <span style="color: magenta; font-weight: 600;">in</span> plan_context.plan.description.lower() <span style="color: magenta; font-weight: 600;">for</span> f <span style="color: magenta; font-weight: 600;">in</span> fragments: base_score = f.relevance_score * <span style="color: yellow;">0.4</span> + f.hierarchy_weight * <span style="color: yellow;">0.3</span> <span style="color: magenta; font-weight: 600;">if</span> is_deploy <span style="color: magenta; font-weight: 600;">and</span> <span style="color: #66cc66;">"uko-infra:"</span> <span style="color: magenta; font-weight: 600;">in</span> f.uko_uri: base_score *= <span style="color: yellow;">1.5</span> <span style="color: #888;"># Boost infrastructure fragments</span> scored.append(ScoredFragment(**vars(f), composite_score=base_score)) <span style="color: magenta; font-weight: 600;">return</span> sorted(scored, key=<span style="color: magenta; font-weight: 600;">lambda</span> s: s.composite_score, reverse=<span style="color: magenta; font-weight: 600;">True</span>)
Register via TOML config (global scope):
[context.pipeline]
fragment-scorer = "my_extensions.scorers:MyCustomScorer"
Or via project-level context view YAML:
project: local/api-service
view: strategize
pipeline:
fragment-scorer: "my_extensions.scorers:MyCustomScorer"
The following table shows which Protocol each pipeline slot implements and what the default implementation provides:
| Pipeline Slot | Protocol | Default Implementation | Design Pattern |
|---|---|---|---|
strategy-selector |
StrategySelectorProtocol |
ConfidenceWeightedSelector |
Strategy |
budget-allocator |
BudgetAllocatorProtocol |
ProportionalBudgetAllocator |
Strategy |
strategy-executor |
StrategyExecutorProtocol |
ParallelStrategyExecutor |
Circuit Breaker |
fragment-deduplicator |
FragmentDeduplicatorProtocol |
ContentHashDeduplicator |
Strategy |
detail-depth-resolver |
DetailDepthResolverProtocol |
MaxDepthResolver |
Specification |
fragment-scorer |
FragmentScorerProtocol |
WeightedCompositeScorer |
Decorator |
budget-packer |
BudgetPackerProtocol |
GreedyKnapsackPacker |
Iterator |
fragment-orderer |
FragmentOrdererProtocol |
RelevanceCoherenceOrderer |
Template Method |
preamble-generator |
PreambleGeneratorProtocol |
ProvenancePreambleGenerator |
Builder |
skeleton-compressor |
SkeletonCompressorProtocol |
DepthReductionCompressor |
Visitor |
Extension Points Summary
| Extension Point | Registration Mechanism | Configuration Format | Discovery |
|---|---|---|---|
| Custom Tools | agents tool add --config <file> |
YAML | Manual registration |
| MCP Server Tools | Referenced in skill YAML mcp_servers section |
YAML | Auto-discovered from MCP server |
| Agent Skills Standard | Referenced in skill YAML agent_skills_dirs section |
Folder structure | Auto-discovered from directory |
| Custom Skills | agents skill add --config <file> |
YAML | Manual registration |
| Custom Actors | agents actor add --config <file> |
YAML | Manual registration |
| Custom Resource Types | agents resource type add --config <file> |
YAML | Manual registration |
| Custom LLM Providers | Python package installation + ProviderRegistry auto-discovery | Python module | Auto-discovered via langchain-* package naming |
| Custom Index Backends | config.toml index.*.custom_module |
TOML + Python module | Configuration-driven |
| Custom Sandbox Strategies | Resource type sandbox_strategy field |
YAML + Python module | Configuration-driven |
| Custom Automation Profiles | agents automation-profile add --config <file> |
YAML | Manual registration |
| Custom Invariant Actors | Standard actor registration + assignment via --invariant-actor |
YAML | Manual assignment |
| Custom UKO Analyzers | config.toml context.uko.analyzers.custom.* |
TOML + Python module | Configuration-driven |
| Custom UKO Vocabularies | Turtle/RDF files + analyzer registration | RDF/Turtle + TOML | Configuration-driven |
| Custom Context Strategies | config.toml context.strategies.custom.* |
TOML + Python module | Configuration-driven |
| Pipeline: StrategySelector | config.toml context.pipeline.strategy-selector or project/plan YAML |
TOML/YAML + Python module | Configuration-driven (scope chain: plan > project > global > built-in) |
| Pipeline: BudgetAllocator | config.toml context.pipeline.budget-allocator or project/plan YAML |
TOML/YAML + Python module | Configuration-driven (scope chain) |
| Pipeline: StrategyExecutor | config.toml context.pipeline.strategy-executor or project/plan YAML |
TOML/YAML + Python module | Configuration-driven (scope chain) |
| Pipeline: FragmentDeduplicator | config.toml context.pipeline.fragment-deduplicator or project/plan YAML |
TOML/YAML + Python module | Configuration-driven (scope chain) |
| Pipeline: DetailDepthResolver | config.toml context.pipeline.detail-depth-resolver or project/plan YAML |
TOML/YAML + Python module | Configuration-driven (scope chain) |
| Pipeline: FragmentScorer | config.toml context.pipeline.fragment-scorer or project/plan YAML |
TOML/YAML + Python module | Configuration-driven (scope chain) |
| Pipeline: BudgetPacker | config.toml context.pipeline.budget-packer or project/plan YAML |
TOML/YAML + Python module | Configuration-driven (scope chain) |
| Pipeline: FragmentOrderer | config.toml context.pipeline.fragment-orderer or project/plan YAML |
TOML/YAML + Python module | Configuration-driven (scope chain) |
| Pipeline: PreambleGenerator | config.toml context.pipeline.preamble-generator or project/plan YAML |
TOML/YAML + Python module | Configuration-driven (scope chain) |
| Pipeline: SkeletonCompressor | config.toml context.pipeline.skeleton-compressor or project/plan YAML |
TOML/YAML + Python module | Configuration-driven (scope chain) |