Replace hard-coded '11 decision types' with dynamic len(DecisionType) in
the assertion and remove the literal from scenario names, step patterns,
docstrings, and docs so the tests stay correct when new DecisionType
values are added.
- Replace dead after_scenario() hook with context.add_cleanup() calls
so patchers are properly stopped (Behave only runs hooks from
environment.py)
- Fix invalid ULID constants: pad DECISION/CORRECTION to 26 chars,
replace invalid 'I' with 'J' in INVARIANT_ULID
- Add temp file cleanup for NamedTemporaryFile(delete=False) via
context.add_cleanup() in behave steps and try/finally in robot helper
- Remove decision-persistence doc sections from testing.md that belong
in PR #438, not this PR
The jitter-spread scenario used millisecond-bucketed wall-clock
timestamps to assert that retried operations did not cluster. On busy
CI runners — especially when a preceding scenario's time.sleep patch
leaked through a late cleanup — tenacity's retry waits became no-ops
and all five operations landed in the same millisecond bucket, tripping
the max_in_bucket <= 3 assertion.
Three hardening changes:
1. Jitter test: switch from time.time() to time.monotonic_ns() and
replace the fragile bucket assertion with a unique-timestamp count
(>= 2 distinct readings among 5 sequential operations).
2. Timeout test: eagerly restore time.sleep in a try/finally block so
subsequent scenarios never observe the patched no-op, regardless of
behave's cleanup ordering.
3. Async circuit breaker: lower recovery_timeout from 0.2 s to 0.1 s
(the sleep step already waits 0.2 s) to give a wider safety margin
on slow CI machines.
Implemented plan-level and project-level locking with configurable timeouts.
Added locks table via Alembic migration storing owner_id, resource_type,
resource_id, acquired_at, and expires_at. Locks enforced in
PlanLifecycleService transitions. Support for re-entrant acquisition,
lock renewal, graceful shutdown release, and startup cleanup of expired
locks. Added diagnostics check for stale lock reporting.
ISSUES CLOSED: #327
- Extract MockMCPTransport to features/mocks/mock_mcp_transport.py
- Use ToolRegistry type with TYPE_CHECKING in register_tools()
- Move transport config validation into __init__ via _validate_config()
Implement MCPToolAdapter to connect to external MCP servers, enumerate
tools, and register them in ToolRegistry with source="mcp". Includes
connect/reconnect/disconnect lifecycle with timeout enforcement, input
validation on invoke, capability inference, Behave/Robot/ASV tests, and
docs/reference/mcp_adapter.md.
ISSUES CLOSED: #159
Extend actor YAML schema to support hierarchical graphs with explicit
node types (agent, tool, conditional, subgraph), per-node LSP bindings
(lsp_binding with server, languages, auto, capabilities), and tool-source
references (skills, mcp_servers, agent_skills).
Add schema validation for namespaced actor references, duplicate node IDs,
edge target existence, and graph reachability — all nodes must be reachable
from entry_node via explicit edges or conditional node routing targets.
Update loader to report YAML parse errors with precise line/column positions
and schema validation errors with dotted field paths and remediation hints
pointing to docs/reference/actor_config.md.
Add docs/reference/actor_config.md as the practical configuration reference
covering hierarchical graph examples, node type table, topology rules, and
common error cases with fix guidance.
Refresh examples/actors/graph_workflow.yaml to replace deprecated actor_path
with actor_ref. Add benchmarks/actor_yaml_bench.py for schema load overhead.
Tests: 95 Behave scenarios, 10 Robot smoke tests (including hierarchical
loader smoke test), security scan clean, coverage 99% (threshold 97%).
ISSUES CLOSED: #157
- Added CHANGELOG.md entry for M2 E2E test suite (CONTRIBUTING.md §6)
- Added Brent E. Edwards to CONTRIBUTORS.md (CONTRIBUTING.md §8)
- Fixed docs/development/testing.md: corrected M2 Behave scenario count
from 11 to 10 to match actual feature file
- Added standard explanatory comments to benchmark sys.path setup
for consistency with project convention (advisory)
- Replaced untyped lambda handler with typed _m2_noop_handler function
in step definitions (advisory)
All nox stages pass: lint, typecheck, format, unit_tests (280 features /
5846 scenarios / 0 failures), integration_tests (682 passed),
coverage_report (97.2%).
ISSUES CLOSED: #169
Added comprehensive test coverage for CLI extension features from #325:
- Behave scenarios for automation_profile resolution, invariant ordering,
and actor override error cases
- Output snapshot assertions for JSON/YAML/table formats
- Robot integration test for action show with optional actors/invariants
- ASV benchmark for extended CLI scenario runtime baseline
- Updated testing.md with CLI extension test fixture documentation
ISSUES CLOSED: #326
Align actor add/remove/list/show commands to YAML-first configs,
namespaced names, and --format json|yaml|plain output support.
Changes:
- Add --format option to list, show, add, and update commands
- Add --update flag to add command for updating existing actors
- Add _actor_spec_dict helper for structured output serialization
- Update _print_actor to support format_output rendering
- Add Behave feature tests for CLI format scenarios (11 scenarios)
- Add Robot Framework integration test for show output fields
- Add ASV benchmark for actor CLI parsing overhead
- Add CLI reference documentation for actor commands
ISSUES CLOSED: #288
The step_register_skills_table step called add_skill in a loop but only
committed once at the end. Because SkillRepository.create() obtains a
new session per call and only flushes (never commits), the intermediate
sessions could be garbage-collected before the final commit, rolling
back their transactions on the shared SQLite :memory: connection. Moving
_commit_pending inside the loop ensures each skill is durably committed
before the next session is created.
ISSUES CLOSED: #418
Add extended CLI flags and output rendering for plan and action
commands in the v3 plan lifecycle.
Plan use command:
- --automation-profile flag validates against BUILTIN_PROFILES and
persists to plan metadata with PLAN provenance
- --invariant flag (repeatable) passes PlanInvariant objects with
InvariantSource.PLAN to PlanLifecycleService.use_action()
- --strategy-actor, --execution-actor, --estimation-actor, and
--invariant-actor flags validate namespace/name format via
validate_namespaced_actor() before setting on the plan
Plan status/list output:
- Profile and Invariants columns added to plan status summary table
- Profile and Invariants columns added to lifecycle-list table
- _plan_spec_dict includes estimation_actor, invariant_actor, and
invariants fields in all output formats (json, yaml, plain, table)
- _print_lifecycle_plan rich panel displays invariants with source
provenance tags
Action show output:
- _print_action displays estimation_actor, invariant_actor,
invariants list, inputs_schema, and automation_profile when present
- _action_spec_dict conditionally includes estimation_actor,
invariant_actor, and inputs_schema fields
Documentation:
- docs/reference/plan_cli.md documents all extended flags, actor
override validation, and usage examples
- docs/reference/action_cli.md documents extended output fields and
YAML configuration options
Tests:
- 29 Behave scenarios in features/cli_extensions.feature covering
automation profile, invariant flags, actor override validation,
plan status/list rendering, and action show extended fields
- 5 Robot Framework integration tests in robot/cli_extensions.robot
for plan use with invariants, profiles, and actor validation
- 6 ASV benchmark suites in benchmarks/cli_extensions_bench.py
measuring parsing overhead for all new flags and rendering paths
ISSUES CLOSED: #325
Each behave-parallel worker previously resolved to the same file-based
SQLite database (cleveragents.db or .cleveragents/db.sqlite) because
before_scenario removed the CLEVERAGENTS_DATABASE_URL env var and the
fallback paths are shared across processes. Under parallel execution
this caused intermittent 'database is locked' and duplicate-project
errors—most visibly in the coverage_report nox session.
Replace the env-var removal in before_scenario with unique temp database
paths via tempfile.mktemp(), giving every scenario its own isolated
database file. Temp files are cleaned up in after_scenario.
Also fix cli_streaming_steps.py where a Background + duplicate Given
sequence triggered a duplicate project name collision, and update
coverage_boost_steps.py so Settings-default assertions explicitly clear
the env vars before testing pydantic defaults.
Reverts the StaticPool fix (68bc068) as it is no longer needed on
develop-hamza-2. Restores the original SingletonThreadPool default
for sqlite:///:memory: engine creation.
The default SingletonThreadPool for sqlite:///:memory: can
non-deterministically lose flushed data when multiple sessions are
created without closing (as in the bulk-register table step).
Under CI with coverage --parallel-mode and behave-parallel --processes 32,
this caused list_all() to return 2 skills instead of 3.
StaticPool guarantees all sessions share the exact same DBAPI connection,
which is the SQLAlchemy-recommended pattern for :memory: test databases.
- extract BaseResourceHandler to eliminate ~90% duplication between
GitCheckoutHandler and FsDirectoryHandler
- raise RuntimeError when sandbox.context is None instead of silent
empty string fallback
- add threading.Lock to resolver handler cache for thread safety
- type resource_lookup/type_lookup as Callable instead of Any
- log original HandlerResolutionError at DEBUG before fallback
- use behave.runner.Context in step definitions per repo convention
- add with_superseded_by() helper for frozen model mutation
- document frozen model + superseded_by interaction in docstring
- clarify sequence_number uniqueness is a persistence concern
- add scenario for invalid corrects_decision_id ULID validation
- add scenario for with_superseded_by copy behavior
- type step helper dict as dict[str, Any]
brent.edwards
freemo
aditya
CoreRasurae
khyari hamza