ci: stop master workflow on PR updates

Remove the stale pull_request trigger from master.yml so PR branch commits do not launch the master workflow.

Maintenance patch for PR #10994.

.forgejo/workflows/master.yml

@@ -3,8 +3,6 @@ name: CI
 on:
     push:
         branches: [master, develop]
-    pull_request:
-        branches: [master, develop]
 
 vars:
     docker_prefix: "http://harbor.cleverthis.com/docker/"

.opencode/agents/implementation-supervisor.md

@@ -245,6 +245,16 @@ each work group's fetch algorithm:
 The prompt body to pass to workers you spawn:
   ```
     Implement or fix the indicated issue or pull request.
+
+    PR Compliance Checklist (MANDATORY — complete ALL items before creating a PR):
+    [ ] 1. CHANGELOG.md — add entry under [Unreleased] section
+    [ ] 2. CONTRIBUTORS.md — add or update contribution entry
+    [ ] 3. Commit footer — include `ISSUES CLOSED: #<issue-number>` in the commit message
+    [ ] 4. CI passes — all quality gates and tests green before requesting review
+    [ ] 5. BDD/Behave tests — added or updated for the changed behaviour
+    [ ] 6. Epic reference — PR description references the parent Epic issue number
+    [ ] 7. Labels — applied via forgejo-label-manager: State/In Review, Priority/<level>, MoSCoW/<level>, Type/<type>
+    [ ] 8. Milestone — PR assigned to the earliest open milestone matching the issue
   ```
 ```
 

CHANGELOG.md

@@ -5,6 +5,32 @@ The format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
 
 ## [Unreleased]
 
+### Added
+
+- **SQLite session persistence and multi-session tab bar with state indicators** (per issue #5330): Added `SessionStore` as a thread-safe, `check_same_thread=False` SQLite-backed persistence layer for TUI sessions, supporting create, retrieve, list, update-state, and delete operations. Added `SessionTabBar` widget that renders named session tabs in the TUI with per-session state indicators (⌛ for working, ❯ for awaiting_input, none for idle) and bracketed highlighting of the active session. The tab bar stays hidden when only one or zero sessions exist. Full BDD test coverage via ``features/tui_session_persistence_tabs.feature`` with 8 scenarios.
+
+### Fixed
+
+- **Cross-actor subgraph cycle detection reads actor_ref field** (#1431): Fixed
+  `_detect_subgraph_cycles()`, `_map_node()`, and the `compile_actor()` main loop
+  in `src/cleveragents/actor/compiler.py` to read `actor_ref` from the top-level
+  `NodeDefinition.actor_ref` field instead of `node.config.get("actor_ref", "")`.
+  Because `actor_ref` is a typed, validated Pydantic field (not a key inside the
+  untyped `config` dict), the old code always returned an empty string, causing
+  cross-actor cycle detection to silently fail and leaving the system vulnerable to
+  infinite recursion at runtime. Added Behave regression tests
+  (`features/actor_subgraph_cycle_detection.feature`) and a Robot Framework
+  integration test (`robot/actor_compiler.robot`) to prevent regressions.
+
+### Changed
+
+- **`agents session list` now displays full 26-character session ULIDs** (#10970): The Rich table
+  and Summary panel ("Most Recent" / "Oldest") previously showed only the first 8 characters of
+  each session ULID. This made the output unusable for copy-paste into `session tell`,
+  `session show`, `session delete`, and `session export`, all of which require the full
+  26-character identifier. The full ULID is now displayed in all output formats (Rich, plain,
+  JSON, YAML, table).
+
 ### Security
 
 - **aiohttp upgraded to >=3.13.4 to remediate CVE-2026-34513 and CVE-2026-34515** (#1549, #1544):
@@ -15,6 +41,16 @@ The format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
   versions (<3.13.4) cannot be installed even if upstream transitive dependencies have loose
   version constraints.
 
+### Fixed
+
+- **Implementation Supervisor PR Compliance Checklist** (#9824): Added a mandatory
+  8-item PR Compliance Checklist to the worker prompt body in `implementation-supervisor.md`
+  that every implementation worker must complete before creating a PR. Checklist covers:
+  CHANGELOG.md update, CONTRIBUTORS.md update, commit footer (`ISSUES CLOSED: #N`),
+  CI verification, BDD tests, Epic reference, label application via `forgejo-label-manager`,
+  and milestone assignment. This eliminates systemic PR merge blockers caused by workers
+  omitting required items.
+
 ### Changed
 
 - Restored `benchmark-regression` CI job to `master.yml` with `pull_request` trigger guard
@@ -79,6 +115,19 @@ The format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
 
 ### Fixed
 
+- **`LLMTraceRepository.save()` premature commit breaks UnitOfWork transactions** (#7505):
+  Replaced the unconditional `session.commit()` in `LLMTraceRepository.save()` with a
+  dual-path implementation that respects the UnitOfWork (UoW) pattern. When an external
+  session is provided (UoW mode), the method now calls only `session.flush()`, leaving
+  transaction control to the caller. When no session is provided (standalone mode), the
+  method creates its own session, flushes, commits, and closes it to ensure durable
+  persistence. This eliminates three data-integrity violations: premature commit of outer
+  UoW transactions, loss of rollback capability for subsequent failures, and a mismatch
+  between the class docstring ("Callers are responsible for commit") and the implementation.
+  Input validation for the `trace` argument was also added. Two new BDD scenarios verify
+  the session contract: `Repository save() calls flush not commit` and `LLM trace rolled
+  back when UnitOfWork transaction rolls back`.
+
 - **git_tools._get_base_env() TOCTOU Race Condition** (#7619): Fixed a
   Time-Of-Check-To-Time-Of-Use race condition in `git_tools._get_base_env()`
   where two concurrent threads could both observe `_BASE_ENV is None`, both
@@ -265,6 +314,14 @@ The format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
 
 ### Added
 
+- **ACMS Index Data Model and File Traversal Engine** (#9579): Implements the
+  foundational ACMS index data model with structured fields for file metadata
+  (path, size, last modified, type), tag system, and hot/warm/cold/archive
+  storage tier assignment. Introduces a timeout-safe large-project file traversal
+  engine capable of handling 10,000+ files without memory exhaustion through
+  chunked processing. Provides a complete index entry pipeline for creation,
+  storage, and retrieval with full queryability by path, tag, type, and recency.
+
 - **ACMS Large-Project Indexing BDD Coverage** (#8726): Added 7 Behave scenarios
   covering walk-based indexing of 10,000+ files without timeout, binary-file
   skipping, oversized-file skipping, git-checkout indexing, fallback to walk when
@@ -273,6 +330,7 @@ The format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
   syscalls). Added `timeout=120` to git subprocess calls to prevent CI hangs.
   Cached `get_scoped_view` results in `When` steps to avoid redundant re-queries
   in `Then` steps.
+
 - **Agent Evolution Pool Supervisor PR Metadata Assignment** (#7888): The
   agent-evolution-pool-supervisor now automatically looks up the Type/Automation
   label and the earliest open milestone from the repository before dispatching
@@ -579,6 +637,15 @@ The format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
   response format from the OpenCode API `/session/status` endpoint instead of an array.
   Workers now dispatch and verify correctly, preventing incorrect session deletion.
 
+---
+### Fixed
+
+- **CLI (`agents actor remove`)** (#6491): Restores output parity with the
+  other actor commands by honoring `--format`/`-f` for JSON/YAML/plain/Rich
+  envelopes. Adds a Robot Framework regression test to assert the JSON
+  envelope structure and updates the CLI synopsis in `docs/specification.md`
+  to document the option.
+
 ---
 
 ## [3.8.0] -- 2026-04-05

CONTRIBUTORS.md

@@ -7,7 +7,6 @@
 * Jeffrey Phillips Freeman <jeffrey.freeman@syncleus.com>
 * Luis Mendes <luis.p.mendes@gmail.com>
 * Rui Hu <rui.hu@cleverthis.com>
-* HAL 9000 <hal9000@cleverthis.com>
 
 # Details
 
@@ -28,4 +27,8 @@ Below are some of the specific details of various contributions.
 * HAL 9000 has contributed the architecture-pool-supervisor milestone assignment feature (PR #8188 / issue #7521): added `forgejo_update_pull_request` permission and documented the PR workflow for major spec changes, enabling automatic milestone assignment for specification PRs.
 * HAL 9000 has contributed the git worktree TOCTOU race condition fix (PR #8178 / issue #7507): replaced the unsafe mkdtemp() + rmdir() pattern with a parent-directory approach to eliminate the race window in concurrent git worktree operations.
 * HAL 9000 has contributed the git_tools TOCTOU race condition fix (PR #8255 / issue #7619): eliminated the Time-Of-Check-To-Time-Of-Use race in `_get_base_env()` by adding double-checked locking with a module-level `threading.Lock`, preventing concurrent threads from writing conflicting environment snapshots.
+* HAL 9000 has contributed the mandatory PR compliance checklist to `implementation-supervisor.md` (#9824): added an 8-item checklist to the worker prompt body with concrete items covering CHANGELOG.md, CONTRIBUTORS.md, commit footer, CI verification, BDD tests, Epic reference, labels, and milestone assignment to eliminate systemic PR merge blockers.
 * HAL 9000 has contributed comprehensive milestone documentation for v3.6.0 (Advanced Concepts & Deferred Features) and v3.7.0 (TUI Implementation) (PR #9903): split into sub-documents covering context strategies, LLM backends, resource types, A2A rename, container tool execution, scope chain resolution, cost/safety budgets, E2E workflow tests, code review examples, plugin architecture, TUI layout, persona system, reference/command input, session management, configuration, and TuiMaterializer integration.
+* HAL 9000 has contributed the LLMTraceRepository data-integrity fix (PR #8185 / issue #7505): replaced the unconditional `session.commit()` in `LLMTraceRepository.save()` with a dual-path implementation that respects the UnitOfWork pattern — flushing only when an external session is provided, and flushing + committing + closing when operating standalone. This eliminates premature transaction commits, loss of rollback capability, and a docstring/implementation mismatch.
+* HAL 9000 has contributed the ACMS Index Data Model and File Traversal Engine (PR #9664 / issue #9579): foundational data structures for indexed context entries with hot/warm/cold/archive storage tier classification, tag system, and a timeout-safe chunked file traversal engine for large projects with 10,000+ files.
+* HAL 9000 has contributed the TUI SQLite session persistence layer and multi-session tab bar widget (per issue #5330): thread-safe `SessionStore` class backed by SQLite with `check_same_thread=False`, proper connection handling (`finally: conn.close()`), ``threading.Lock`` protection around all mutating operations, replacement of deprecated ``datetime.utcnow()`` with ``datetime.now(timezone.utc)``, and state-aware session tab rendering (⌛ working, ❯ awaiting_input, none idle) via the `SessionTabBar` Textual widget. Includes 8 BDD scenarios.

docs/reference/session_cli.md

@@ -72,8 +72,8 @@ The `rich` format renders a sessions table with columns: **ID**, **Name**, **Act
 | Field | Description |
 |-------|-------------|
 | Total | Number of sessions |
-| Most Recent | Name or truncated ID of the most recently updated session |
-| Oldest | Name or truncated ID of the oldest session |
+| Most Recent | Name or full ULID of the most recently updated session |
+| Oldest | Name or full ULID of the oldest session |
 | Total Messages | Sum of messages across all sessions |
 | Storage | Estimated storage used |
 
@@ -85,7 +85,7 @@ Followed by a `✓ OK N sessions listed` success message.
 {
   "sessions": [
     {
-      "id": "01HXYZ...",
+      "id": "01HXYZ4M1Q3F0R0E5HR8K5T8A",
       "name": "my-session",
       "actor": "openai/gpt-4",
       "messages": 5,

docs/showcase/cli-tools/project-init-and-context-management.md

@@ -296,13 +296,13 @@ $ python -m cleveragents session list
 ┏━━━━━━━━━━┳━━━━━━━━━━━┳━━━━━━━━┳━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━┓
 ┃ ID       ┃ Name      ┃ Actor  ┃ Messages ┃ Updated          ┃
 ┡━━━━━━━━━━╇━━━━━━━━━━━╇━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━┩
-│ 01KNKK4Q │ (unnamed) │ (none) │        0 │ 2026-04-07 09:07 │
+│ 01KNKK4Q9GZ0TRR5B0NEJYGMWH │ (unnamed) │ (none) │        0 │ 2026-04-07 09:07 │
 └──────────┴───────────┴────────┴──────────┴──────────────────┘
 
 ╭────────────────────────────────── Summary ───────────────────────────────────╮
 │ Total:          1                                                            │
-│ Most Recent:    01KNKK4Q                                                     │
-│ Oldest:         01KNKK4Q                                                     │
+│ Most Recent:    01KNKK4Q9GZ0TRR5B0NEJYGMWH                                                     │
+│ Oldest:         01KNKK4Q9GZ0TRR5B0NEJYGMWH                                                     │
 │ Total Messages: 0                                                            │
 │ Storage:        0 KB                                                         │
 ╰──────────────────────────────────────────────────────────────────────────────╯
@@ -311,7 +311,7 @@ $ python -m cleveragents session list
 ```
 
 **What's Happening:**
-The session list shows all sessions with their truncated ID, optional name, bound actor, message count, and last update time. The summary panel provides aggregate statistics across all sessions.
+The session list shows all sessions with their full ULID, optional name, bound actor, message count, and last update time. The summary panel provides aggregate statistics across all sessions.
 
 ---
 
@@ -339,8 +339,8 @@ $ python -m cleveragents session list --format json
     ],
     "summary": {
       "total": 1,
-      "most_recent": "01KNKK4Q",
-      "oldest": "01KNKK4Q",
+      "most_recent": "01KNKK4Q9GZ0TRR5B0NEJYGMWH",
+      "oldest": "01KNKK4Q9GZ0TRR5B0NEJYGMWH",
       "total_messages": 0,
       "storage": "0 KB"
     }
@@ -469,7 +469,7 @@ $ python -m cleveragents session list
 ┏━━━━━━━━━━┳━━━━━━━━━━━┳━━━━━━━━┳━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━┓
 ┃ ID       ┃ Name      ┃ Actor  ┃ Messages ┃ Updated          ┃
 ┡━━━━━━━━━━╇━━━━━━━━━━━╇━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━┩
-│ 01KNKK4Q │ (unnamed) │ (none) │        0 │ 2026-04-07 09:07 │
+│ 01KNKK4Q9GZ0TRR5B0NEJYGMWH │ (unnamed) │ (none) │        0 │ 2026-04-07 09:07 │
 └──────────┴───────────┴────────┴──────────┴──────────────────┘
 ✓ OK 1 sessions listed
 ```

docs/showcase/cli-tools/session-management-workflows.md

@@ -180,14 +180,14 @@ $ python -m cleveragents session list
 ┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
 ┃ ID       ┃ Name       ┃ Actor          ┃ Messages ┃ Updated          ┃
 ┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩
-│ 01HXYZ4M │ (unnamed)  │ openai/gpt-4o  │        3 │ 2026-04-07 09:22 │
-│ 01HXYZ3K │ (unnamed)  │ (none)         │        0 │ 2026-04-07 09:15 │
+│ 01HXYZ4M1Q3F0R0E5HR8K5T8A │ (unnamed)  │ openai/gpt-4o  │        3 │ 2026-04-07 09:22 │
+│ 01HXYZ3K9P2E9Q9D4GQ7J4S7Z │ (unnamed)  │ (none)         │        0 │ 2026-04-07 09:15 │
 └──────────┴────────────┴────────────────┴──────────┴──────────────────┘
 
 ╭──────────────────────────────────────────── Summary ─────────────────────────────────────────────╮
 │ Total:          2                                                                                 │
-│ Most Recent:    01HXYZ4M                                                                          │
-│ Oldest:         01HXYZ3K                                                                          │
+│ Most Recent:    01HXYZ4M1Q3F0R0E5HR8K5T8A                                                                          │
+│ Oldest:         01HXYZ3K9P2E9Q9D4GQ7J4S7Z                                                                          │
 │ Total Messages: 3                                                                                 │
 │ Storage:        0 KB                                                                              │
 ╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
@@ -196,8 +196,8 @@ $ python -m cleveragents session list
 ```
 
 **What's Happening:**
-The list command renders a Rich table with five columns: truncated ID (first
-8 characters for readability), optional name, bound actor, message count, and
+The list command renders a Rich table with five columns: the full session
+ULID (26 characters), optional name, bound actor, message count, and
 last update time. The **Summary** panel below shows aggregate statistics
 including total sessions, most recent, oldest, total message count, and
 storage used.
@@ -229,8 +229,8 @@ $ python -m cleveragents session list --format json
   ],
   "summary": {
     "total": 2,
-    "most_recent": "01HXYZ4M",
-    "oldest": "01HXYZ3K",
+    "most_recent": "01HXYZ4M1Q3F0R0E5HR8K5T8A",
+    "oldest": "01HXYZ3K9P2E9Q9D4GQ7J4S7Z",
     "total_messages": 3,
     "storage": "0 KB"
   }
@@ -773,7 +773,7 @@ $ python -m cleveragents session list
 ✓ OK 2 sessions listed
 
 $ python -m cleveragents session list --format json
-{"sessions": [...], "summary": {"total": 2, "most_recent": "01HXYZ4M", ...}}
+{"sessions": [...], "summary": {"total": 2, "most_recent": "01HXYZ4M1Q3F0R0E5HR8K5T8A", ...}}
 
 $ python -m cleveragents session tell --session 01HXYZ4M1Q3F0R0E5HR8K5T8A "What is the capital of France?"
 user: What is the capital of France?

docs/specification.md

@@ -277,7 +277,7 @@ The following standards are integrated into the architecture:
                  [(<span style="color: magenta;"><span style="color: cyan;">--temperature</span>|<span style="color: yellow;">-t</span></span>) <span style="color: #66cc66;">&lt;TEMP&gt;</span>] [<span style="color: cyan;">--allow-rxpy-in-run-mode</span>]
                  [<span style="color: cyan;">--skill</span> <span style="color: #66cc66;">&lt;SKILL&gt;</span>]... <span style="color: #66cc66;">&lt;NAME&gt;</span> <span style="color: #66cc66;">&lt;PROMPT&gt;</span>
 <span style="color: cyan; font-weight: 600;">agents</span> actor add <span style="color: cyan;">--config</span>|<span style="color: yellow;">-c</span> <span style="color: #66cc66;">&lt;FILE&gt;</span> [<span style="color: cyan;">--update</span>]
-<span style="color: cyan; font-weight: 600;">agents</span> actor remove <span style="color: #66cc66;">&lt;NAME&gt;</span>
+<span style="color: cyan; font-weight: 600;">agents</span> actor remove [<span style="color: cyan;">--format</span> <span style="color: #66cc66;">&lt;FORMAT&gt;</span>] <span style="color: #66cc66;">&lt;NAME&gt;</span>
 <span style="color: cyan; font-weight: 600;">agents</span> actor list
 <span style="color: cyan; font-weight: 600;">agents</span> actor show <span style="color: #66cc66;">&lt;NAME&gt;</span>
 <span style="color: cyan; font-weight: 600;">agents</span> actor context remove [<span style="color: cyan;">--yes</span>|<span style="color: yellow;">-y</span>] (<span style="color: magenta;"><span style="color: cyan;">--all</span>|<span style="color: yellow;">-a</span>|<span style="color: #66cc66;">&lt;NAME&gt;</span></span>)
@@ -1723,8 +1723,8 @@ None.
     ╭─ Sessions ───────────────────────────────────────────────────────────────────╮
     │ <span style="color: cyan; font-weight: 600;">ID</span>        <span style="color: cyan; font-weight: 600;">Name</span>             <span style="color: cyan; font-weight: 600;">Actor</span>               <span style="color: cyan; font-weight: 600;">Messages</span>  <span style="color: cyan; font-weight: 600;">Updated</span>             │
     │ <span style="opacity: 0.7;">────────  ───────────────  ──────────────────  ────────  ────────────────</span>    │
-    │ 01HXM2A6  weekly-planning  local/orchestrator  6         2026-02-08 12:44    │
-    │ 01HXM1F2  refactor-sprint  local/orchestrator  14        2026-02-07 18:11    │
+    │ 01HXM2A61MQHZ4MRBAY3MPNJTN  weekly-planning  local/orchestrator  6         2026-02-08 12:44    │
+    │ 01HXM1F21MQHZ4MRBAY3MPNJTN  refactor-sprint  local/orchestrator  14        2026-02-07 18:11    │
     ╰──────────────────────────────────────────────────────────────────────────────╯
 
     ╭─ Summary ────────────────────╮
@@ -1746,8 +1746,8 @@ None.
     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
+    01HXM2A61MQHZ4MRBAY3MPNJTN  weekly-planning  local/orchestrator  6         2026-02-08 12:44
+    01HXM1F21MQHZ4MRBAY3MPNJTN  refactor-sprint  local/orchestrator  14        2026-02-07 18:11
 
     Summary
       Total: 2
@@ -1769,14 +1769,14 @@ None.
       "data": {
         "sessions": [
           {
-            "id": "01HXM2A6",
+            "id": "01HXM2A61MQHZ4MRBAY3MPNJTN",
             "name": "weekly-planning",
             "actor": "local/orchestrator",
             "messages": 6,
             "updated": "2026-02-08T12:44:00Z"
           },
           {
-            "id": "01HXM1F2",
+            "id": "01HXM1F21MQHZ4MRBAY3MPNJTN",
             "name": "refactor-sprint",
             "actor": "local/orchestrator",
             "messages": 14,
@@ -1804,12 +1804,12 @@ None.
     exit_code: 0
     data:
       sessions:
-        - id: 01HXM2A6
+        - id: 01HXM2A61MQHZ4MRBAY3MPNJTN
           name: weekly-planning
           actor: local/orchestrator
           messages: 6
           updated: "2026-02-08T12:44:00Z"
-        - id: 01HXM1F2
+        - id: 01HXM1F21MQHZ4MRBAY3MPNJTN
           name: refactor-sprint
           actor: local/orchestrator
           messages: 14

features/acms/index_data_model_and_traversal.feature

@@ -0,0 +1,138 @@
+Feature: ACMS Index Data Model and File Traversal Engine
+  As a developer
+  I want to index large projects with 10,000+ files
+  So that I can efficiently query and manage context entries at scale
+
+  Background:
+    Given I have an ACMS index
+    And I have a file traversal engine with chunk size 100
+
+  Scenario: Create an index entry with file metadata
+    When I create an index entry with:
+      | path | /project/src/main.py |
+      | file_type | python |
+      | size_bytes | 1024 |
+    Then the index entry should have path "/project/src/main.py"
+    And the index entry should have file type "python"
+    And the index entry should have size 1024 bytes
+
+  Scenario: Add tags to an index entry
+    Given I have an index entry with path "/project/src/main.py"
+    When I add tag "core" to the entry
+    And I add tag "important" to the entry
+    Then the entry should have tag "core"
+    And the entry should have tag "important"
+    And the entry should have 2 tags
+
+  Scenario: Set tier level for an index entry
+    Given I have an index entry with path "/project/src/main.py"
+    When I set the tier level to "hot"
+    Then the entry should have tier level "hot"
+
+  Scenario: Add entry to index
+    Given I have an index entry with path "/project/src/main.py"
+    When I add the entry to the index
+    Then the index should contain 1 entry
+    And I should be able to retrieve the entry by path "/project/src/main.py"
+
+  Scenario: Query index by path pattern
+    Given I have an index with entries:
+      | path | file_type |
+      | /project/src/main.py | python |
+      | /project/src/utils.py | python |
+      | /project/tests/test_main.py | python |
+      | /project/docs/readme.md | markdown |
+    When I query the index by path pattern "src"
+    Then I should get 2 results
+    And the results should include "/project/src/main.py"
+    And the results should include "/project/src/utils.py"
+
+  Scenario: Query index by file type
+    Given I have an index with entries:
+      | path | file_type |
+      | /project/src/main.py | python |
+      | /project/src/utils.py | python |
+      | /project/src/app.js | javascript |
+      | /project/docs/readme.md | markdown |
+    When I query the index by file type "python"
+    Then I should get 2 results
+    And all results should have file type "python"
+
+  Scenario: Query index by tag
+    Given I have an index with entries:
+      | path | tags |
+      | /project/src/main.py | core,important |
+      | /project/src/utils.py | supporting |
+      | /project/tests/test_main.py | test,important |
+    When I query the index by tag "important"
+    Then I should get 2 results
+    And the results should include "/project/src/main.py"
+    And the results should include "/project/tests/test_main.py"
+
+  Scenario: Query index by tier level
+    Given I have an index with entries:
+      | path | tier |
+      | /project/src/main.py | hot |
+      | /project/src/utils.py | warm |
+      | /project/tests/test_main.py | cold |
+    When I query the index by tier level "hot"
+    Then I should get 1 result
+    And the result should have path "/project/src/main.py"
+
+  Scenario: Query index by recency
+    Given I have an index with entries from different dates
+    When I query the index for entries modified after "2026-04-01"
+    Then I should get entries modified after that date
+
+  Scenario: Traverse and index a directory with multiple files
+    Given I have a test directory with 50 files
+    When I traverse and index the directory
+    Then the index should contain 50 entries
+    And all entries should have valid file paths
+
+  Scenario: Handle large project traversal with chunked processing
+    Given I have a test directory with 1000 files
+    When I traverse and index the directory with chunk size 100
+    Then the index should contain 1000 entries
+    And the traversal should complete without timeout
+
+  Scenario: Exclude patterns during traversal
+    Given I have a test directory with files including:
+      | path |
+      | /project/src/main.py |
+      | /project/.git/config |
+      | /project/__pycache__/main.cpython-39.pyc |
+      | /project/src/utils.py |
+    When I traverse and index the directory excluding ".git" and "__pycache__"
+    Then the index should contain 2 entries
+    And the index should not contain ".git" paths
+    And the index should not contain "__pycache__" paths
+
+  Scenario: Get all entries from index
+    Given I have an index with 5 entries
+    When I get all entries from the index
+    Then I should get 5 results
+
+  Scenario: Get entry count from index
+    Given I have an index with 10 entries
+    When I get the entry count
+    Then the count should be 10
+
+  Scenario: Remove entry from index
+    Given I have an index with 3 entries
+    When I remove an entry by path
+    Then the index should contain 2 entries
+
+  Scenario: Combined query with multiple filters
+    Given I have an index with entries:
+      | path | file_type | tags | tier |
+      | /project/src/main.py | python | core,important | hot |
+      | /project/src/utils.py | python | supporting | warm |
+      | /project/tests/test_main.py | python | test | cold |
+      | /project/docs/readme.md | markdown | docs | cold |
+    When I query the index with filters:
+      | path_pattern | src |
+      | file_type | python |
+      | tier | hot |
+    Then I should get 1 result
+    And the result should have path "/project/src/main.py"

features/actor_cli_yaml.feature

@@ -59,6 +59,11 @@ Feature: Actor CLI YAML-first alignment
     When I run actor remove with namespaced name
     Then the actor remove should succeed for namespaced name
 
+  Scenario: Actor remove outputs JSON format
+    Given an actor CLI runner
+    When I run actor remove with format json
+    Then the actor remove output should be valid JSON envelope
+
   Scenario: Actor update outputs JSON format
     Given an actor CLI runner
     When I run actor update with format json

features/actor_subgraph_cycle_detection.feature

@@ -0,0 +1,38 @@
+Feature: Cross-actor subgraph cycle detection reads actor_ref field
+  As a CleverAgents developer
+  I want the actor compiler to correctly detect cross-actor subgraph cycles
+  So that mutually-referencing actors raise SubgraphCycleError at compile time
+
+  Background:
+    Given the actor compiler is available
+
+  # ────────────────────────────────────────────────────────────
+  # Bug fix: actor_ref is a top-level NodeDefinition field
+  # ────────────────────────────────────────────────────────────
+
+  @tdd_issue @tdd_issue_1431
+  Scenario: Mutually-referencing actors raise SubgraphCycleError
+    Given actor "test/actor-a" has a subgraph node with actor_ref "test/actor-b"
+    And actor "test/actor-b" has a subgraph node with actor_ref "test/actor-a"
+    And a registry containing both actors
+    When I compile actor "test/actor-a" with the registry resolver
+    Then the compilation should raise SubgraphCycleError
+    And the cycle error message should mention "cycle"
+
+  @tdd_issue @tdd_issue_1431
+  Scenario: Non-cyclic subgraph reference compiles successfully
+    Given actor "test/actor-x" has a subgraph node with actor_ref "test/actor-y"
+    And actor "test/actor-y" has no subgraph nodes
+    And a registry containing both actors
+    When I compile actor "test/actor-x" with the registry resolver
+    Then the actor compilation should succeed
+    And the actor subgraph_refs should map "sub" to "test/actor-y"
+
+  @tdd_issue @tdd_issue_1431
+  Scenario: Subgraph node actor_ref is reflected in compiled metadata
+    Given actor "test/actor-p" has a subgraph node with actor_ref "test/actor-q"
+    And actor "test/actor-q" has no subgraph nodes
+    And a registry containing both actors
+    When I compile actor "test/actor-p" with the registry resolver
+    Then the actor compilation should succeed
+    And the compiled node "sub" should have subgraph set to "test/actor-q"

features/consolidated_misc.feature

@@ -1714,6 +1714,12 @@ Feature: Consolidated Misc
     And the temporary connection should be closed afterward
 
 
+  Scenario: get_current_revision uses check_same_thread=False for SQLite engines
+    Given a migration runner configured for "sqlite:///:memory:"
+    When I request the current revision from the database
+    Then the SQLite engine for get_current_revision should use check_same_thread=False
+
+
   Scenario: File-based SQLite database directory is created if missing
     Given a migration runner configured for "sqlite:///tmp/test-db/mydb.db"
     When I initialize or upgrade a file-based SQLite database

features/environment.py

@@ -686,6 +686,12 @@ def after_scenario(context, scenario):
             pass  # Ignore cleanup errors
         context.test_dir = None
 
+    # Clean up TemporaryDirectory objects created by ACMS index traversal tests
+    if hasattr(context, "temp_dir") and context.temp_dir is not None:
+        with contextlib.suppress(Exception):
+            context.temp_dir.cleanup()
+        context.temp_dir = None
+
     # Clean up environment variables set during tests
     if hasattr(context, "env_vars_to_clean"):
         for key in context.env_vars_to_clean:

features/pr_compliance_checklist.feature

@@ -0,0 +1,58 @@
+@mock_only
+Feature: PR Compliance Checklist in Implementation Supervisor
+
+  As an implementation supervisor
+  I want to pass a mandatory PR compliance checklist to every worker prompt
+  So that workers complete all required items before creating a PR and avoid systemic merge blockers
+
+  Background:
+    Given the implementation-supervisor.md agent definition exists
+
+  Scenario: Supervisor worker prompt includes the PR compliance checklist
+    When I read the implementation supervisor agent definition
+    Then the worker prompt body includes the PR compliance checklist section
+    And the checklist is marked as MANDATORY
+
+  Scenario: Checklist item 1 — CHANGELOG.md update required
+    When I read the implementation supervisor agent definition
+    Then the worker prompt body includes a CHANGELOG.md checklist item
+    And the item instructs workers to add an entry under the Unreleased section
+
+  Scenario: Checklist item 2 — CONTRIBUTORS.md update required
+    When I read the implementation supervisor agent definition
+    Then the worker prompt body includes a CONTRIBUTORS.md checklist item
+    And the item instructs workers to add or update their contribution entry
+
+  Scenario: Checklist item 3 — commit footer required
+    When I read the implementation supervisor agent definition
+    Then the worker prompt body includes a commit footer checklist item
+    And the item specifies the ISSUES CLOSED footer format
+
+  Scenario: Checklist item 4 — CI must pass before PR creation
+    When I read the implementation supervisor agent definition
+    Then the worker prompt body includes a CI passes checklist item
+    And the item instructs workers to verify all quality gates are green
+
+  Scenario: Checklist item 5 — BDD/Behave tests required
+    When I read the implementation supervisor agent definition
+    Then the worker prompt body includes a BDD tests checklist item
+    And the item instructs workers to add or update Behave feature files
+
+  Scenario: Checklist item 6 — Epic reference required in PR description
+    When I read the implementation supervisor agent definition
+    Then the worker prompt body includes an Epic reference checklist item
+    And the item instructs workers to reference the parent Epic issue number
+
+  Scenario: Checklist item 7 — Labels must be applied
+    When I read the implementation supervisor agent definition
+    Then the worker prompt body includes a labels checklist item
+    And the item instructs workers to apply labels via forgejo-label-manager
+
+  Scenario: Checklist item 8 — Milestone must be assigned
+    When I read the implementation supervisor agent definition
+    Then the worker prompt body includes a milestone checklist item
+    And the item instructs workers to assign the earliest open milestone
+
+  Scenario: All 8 checklist items are present in the worker prompt
+    When I read the implementation supervisor agent definition
+    Then the worker prompt body contains all 8 mandatory checklist items

features/session_cli.feature

@@ -44,6 +44,28 @@ Feature: Session CLI commands
     When I run session CLI list with --format json
     Then the session CLI JSON list entries should match the documented contract
 
+  Scenario: List sessions displays full 26-character ULIDs in Rich table
+    Given there are mocked existing sessions
+    When I run session CLI list
+    Then the session CLI rich table should display full session ULIDs
+
+  Scenario: List sessions summary panel shows full ULIDs for unnamed sessions
+    Given there are mocked existing sessions
+    When I run session CLI list
+    Then the session CLI summary panel should contain full session ULIDs
+
+  Scenario: List sessions summary panel shows session names for named sessions
+    Given there are mocked existing named sessions
+    When I run session CLI list
+    Then the session CLI summary panel should show session names
+
+  Scenario: Full session ID from list output works with session tell
+    Given there are mocked existing sessions
+    When I run session CLI list
+    And I capture the first session full ULID from the output
+    And I run session CLI tell with the full session ID and prompt "Hello from list"
+    Then the session CLI tell should succeed
+
   # Show command tests
   Scenario: Show session with valid ID
     Given there is a mocked session with messages

features/steps/acms_index_data_model_traversal_steps.py

@@ -0,0 +1,395 @@
+"""Step definitions for ACMS Index Data Model and File Traversal Engine tests."""
+
+from __future__ import annotations
+
+import tempfile
+from datetime import datetime, timedelta
+from pathlib import Path
+
+from behave import given, then, when
+
+from cleveragents.acms.index import (
+    ACMSIndex,
+    FileTraversalEngine,
+    FileType,
+    IndexEntry,
+    TierLevel,
+)
+
+
+@given("I have an ACMS index")
+def step_create_index(context):
+    """Create a new ACMS index."""
+    context.index = ACMSIndex()
+
+
+@given("I have a file traversal engine with chunk size {chunk_size:d}")
+def step_create_traversal_engine(context, chunk_size):
+    """Create a file traversal engine with specified chunk size."""
+    context.engine = FileTraversalEngine(chunk_size=chunk_size)
+
+
+@when("I create an index entry with:")
+def step_create_index_entry(context):
+    """Create an index entry from table data."""
+    data = {row["key"]: row["value"] for row in context.table}
+
+    file_type = FileType(data.get("file_type", "other"))
+    size_bytes = int(data.get("size_bytes", "0"))
+
+    context.entry = IndexEntry(
+        path=data["path"],
+        file_type=file_type,
+        size_bytes=size_bytes,
+        created_at=datetime.now(),
+        modified_at=datetime.now(),
+    )
+
+
+@then('the index entry should have path "{path}"')
+def step_check_entry_path(context, path):
+    """Verify the index entry has the expected path."""
+    assert context.entry.path == path
+
+
+@then('the index entry should have file type "{file_type}"')
+def step_check_entry_file_type(context, file_type):
+    """Verify the index entry has the expected file type."""
+    assert context.entry.file_type == FileType(file_type)
+
+
+@then("the index entry should have size {size:d} bytes")
+def step_check_entry_size(context, size):
+    """Verify the index entry has the expected size."""
+    assert context.entry.size_bytes == size
+
+
+@given('I have an index entry with path "{path}"')
+def step_create_entry_with_path(context, path):
+    """Create an index entry with a specific path."""
+    context.entry = IndexEntry(
+        path=path,
+        file_type=FileType.PYTHON,
+        size_bytes=1024,
+        created_at=datetime.now(),
+        modified_at=datetime.now(),
+    )
+
+
+@when('I add tag "{tag}" to the entry')
+def step_add_tag_to_entry(context, tag):
+    """Add a tag to the current entry."""
+    context.entry.add_tag(tag)
+
+
+@then('the entry should have tag "{tag}"')
+def step_check_entry_has_tag(context, tag):
+    """Verify the entry has a specific tag."""
+    assert context.entry.has_tag(tag)
+
+
+@then("the entry should have {count:d} tags")
+def step_check_entry_tag_count(context, count):
+    """Verify the entry has the expected number of tags."""
+    assert len(context.entry.tags) == count
+
+
+@when('I set the tier level to "{tier}"')
+def step_set_entry_tier(context, tier):
+    """Set the tier level for the entry."""
+    context.entry.set_tier(TierLevel(tier))
+
+
+@then('the entry should have tier level "{tier}"')
+def step_check_entry_tier(context, tier):
+    """Verify the entry has the expected tier level."""
+    assert context.entry.tier == TierLevel(tier)
+
+
+@when("I add the entry to the index")
+def step_add_entry_to_index(context):
+    """Add the current entry to the index."""
+    context.index.add_entry(context.entry)
+
+
+@then("the index should contain {count:d} entry")
+def step_check_index_entry_count_singular(context, count):
+    """Verify the index has the expected number of entries."""
+    assert context.index.get_entry_count() == count
+
+
+@then("the index should contain {count:d} entries")
+def step_check_index_entry_count(context, count):
+    """Verify the index has the expected number of entries."""
+    assert context.index.get_entry_count() == count
+
+
+@then('I should be able to retrieve the entry by path "{path}"')
+def step_retrieve_entry_by_path(context, path):
+    """Verify we can retrieve an entry by path."""
+    entry = context.index.get_entry(path)
+    assert entry is not None
+    assert entry.path == path
+
+
+@given("I have an index with entries:")
+def step_create_index_with_entries(context):
+    """Create an index with multiple entries from table data."""
+    context.index = ACMSIndex()
+
+    for row in context.table:
+        path = row["path"]
+        file_type = FileType(row.get("file_type", "other"))
+
+        entry = IndexEntry(
+            path=path,
+            file_type=file_type,
+            size_bytes=1024,
+            created_at=datetime.now(),
+            modified_at=datetime.now(),
+        )
+
+        # Add tags if present
+        if "tags" in row:
+            for tag in row["tags"].split(","):
+                entry.add_tag(tag.strip())
+
+        # Set tier if present
+        if "tier" in row:
+            entry.set_tier(TierLevel(row["tier"]))
+
+        context.index.add_entry(entry)
+
+
+@given("I have an index with {count:d} entries")
+def step_create_index_with_n_entries(context, count):
+    """Create an index with a specified number of entries."""
+    context.index = ACMSIndex()
+    for i in range(count):
+        entry = IndexEntry(
+            path=f"/project/file{i}.py",
+            file_type=FileType.PYTHON,
+            size_bytes=1024,
+            created_at=datetime.now(),
+            modified_at=datetime.now(),
+        )
+        context.index.add_entry(entry)
+
+
+@when('I query the index by path pattern "{pattern}"')
+def step_query_by_path_pattern(context, pattern):
+    """Query the index by path pattern."""
+    context.query_results = context.index.query_by_path(pattern)
+
+
+@then("I should get {count:d} results")
+def step_check_query_result_count(context, count):
+    """Verify the query returned the expected number of results."""
+    assert len(context.query_results) == count, (
+        f"Expected {count} results, got {len(context.query_results)}"
+    )
+
+
+@then("I should get {count:d} result")
+def step_check_query_result_count_singular(context, count):
+    """Verify the query returned the expected number of results (singular)."""
+    assert len(context.query_results) == count, (
+        f"Expected {count} result, got {len(context.query_results)}"
+    )
+
+
+@then('the results should include "{path}"')
+def step_check_result_includes_path(context, path):
+    """Verify the query results include a specific path."""
+    paths = [entry.path for entry in context.query_results]
+    assert path in paths
+
+
+@when('I query the index by file type "{file_type}"')
+def step_query_by_file_type(context, file_type):
+    """Query the index by file type."""
+    context.query_results = context.index.query_by_type(FileType(file_type))
+
+
+@then('all results should have file type "{file_type}"')
+def step_check_all_results_file_type(context, file_type):
+    """Verify all results have the expected file type."""
+    expected_type = FileType(file_type)
+    for entry in context.query_results:
+        assert entry.file_type == expected_type
+
+
+@when('I query the index by tag "{tag}"')
+def step_query_by_tag(context, tag):
+    """Query the index by tag."""
+    context.query_results = context.index.query_by_tag(tag)
+
+
+@when('I query the index by tier level "{tier}"')
+def step_query_by_tier(context, tier):
+    """Query the index by tier level."""
+    context.query_results = context.index.query_by_tier(TierLevel(tier))
+
+
+@then('the result should have path "{path}"')
+def step_check_single_result_path(context, path):
+    """Verify the single result has the expected path."""
+    assert len(context.query_results) == 1
+    assert context.query_results[0].path == path
+
+
+@given("I have an index with entries from different dates")
+def step_create_index_with_dated_entries(context):
+    """Create an index with entries from different dates."""
+    context.index = ACMSIndex()
+
+    now = datetime.now()
+    dates = [
+        now - timedelta(days=10),
+        now - timedelta(days=5),
+        now - timedelta(days=1),
+        now,
+    ]
+
+    for i, date in enumerate(dates):
+        entry = IndexEntry(
+            path=f"/project/file{i}.py",
+            file_type=FileType.PYTHON,
+            size_bytes=1024,
+            created_at=date,
+            modified_at=date,
+        )
+        context.index.add_entry(entry)
+
+
+@when('I query the index for entries modified after "{date_str}"')
+def step_query_by_recency(context, date_str):
+    """Query the index for entries modified after a specific date."""
+    # Parse date string (format: YYYY-MM-DD)
+    date = datetime.strptime(date_str, "%Y-%m-%d")
+    context.query_results = context.index.query_by_recency(date)
+
+
+@then("I should get entries modified after that date")
+def step_check_recency_results(context):
+    """Verify the recency query returned valid results."""
+    assert len(context.query_results) > 0
+
+
+@given("I have a test directory with {count:d} files")
+def step_create_test_directory(context, count):
+    """Create a temporary directory with test files."""
+    context.temp_dir = tempfile.TemporaryDirectory()
+    temp_path = Path(context.temp_dir.name)
+
+    # Create subdirectories and files
+    for i in range(count):
+        subdir = temp_path / f"subdir{i % 10}"
+        subdir.mkdir(exist_ok=True)
+
+        file_path = subdir / f"file{i}.py"
+        file_path.write_text(f"# Test file {i}\nprint('Hello {i}')\n")
+
+
+@when("I traverse and index the directory")
+def step_traverse_and_index(context):
+    """Traverse and index the test directory."""
+    context.engine.reset_index()
+    context.index = context.engine.traverse_and_index(context.temp_dir.name)
+
+
+@when("I traverse and index the directory with chunk size {chunk_size:d}")
+def step_traverse_and_index_with_chunk_size(context, chunk_size):
+    """Traverse and index the test directory with a specific chunk size."""
+    engine = FileTraversalEngine(chunk_size=chunk_size)
+    context.index = engine.traverse_and_index(context.temp_dir.name)
+
+
+@then("all entries should have valid file paths")
+def step_check_valid_file_paths(context):
+    """Verify all entries have valid file paths."""
+    for entry in context.index.get_all_entries():
+        assert entry.path
+        assert len(entry.path) > 0
+
+
+@then("the traversal should complete without timeout")
+def step_check_no_timeout(context):
+    """Verify the traversal completed without timeout."""
+    # This step passes if we got here without timing out
+    assert True
+
+
+@given("I have a test directory with files including:")
+def step_create_test_directory_with_specific_files(context):
+    """Create a test directory with specific files."""
+    context.temp_dir = tempfile.TemporaryDirectory()
+    temp_path = Path(context.temp_dir.name)
+
+    for row in context.table:
+        file_path = temp_path / row["path"].lstrip("/")
+        file_path.parent.mkdir(parents=True, exist_ok=True)
+        file_path.write_text("test content")
+
+
+@when('I traverse and index the directory excluding "{exclude1}" and "{exclude2}"')
+def step_traverse_with_exclusions(context, exclude1, exclude2):
+    """Traverse and index with exclusion patterns."""
+    context.engine.reset_index()
+    context.index = context.engine.traverse_and_index(
+        context.temp_dir.name,
+        exclude_patterns=[exclude1, exclude2],
+    )
+
+
+@then('the index should not contain "{pattern}" paths')
+def step_check_no_excluded_paths(context, pattern):
+    """Verify the index doesn't contain paths matching the pattern."""
+    for entry in context.index.get_all_entries():
+        assert pattern not in entry.path
+
+
+@when("I get all entries from the index")
+def step_get_all_entries(context):
+    """Get all entries from the index."""
+    context.query_results = context.index.get_all_entries()
+
+
+@when("I get the entry count")
+def step_get_entry_count(context):
+    """Get the entry count from the index."""
+    context.entry_count = context.index.get_entry_count()
+
+
+@then("the count should be {count:d}")
+def step_check_entry_count_value(context, count):
+    """Verify the entry count matches the expected value."""
+    assert context.entry_count == count
+
+
+@when("I remove an entry by path")
+def step_remove_entry(context):
+    """Remove an entry from the index."""
+    # Get the first entry's path
+    entries = context.index.get_all_entries()
+    if entries:
+        context.index.remove_entry(entries[0].path)
+
+
+@when("I query the index with filters:")
+def step_query_with_combined_filters(context):
+    """Query the index with multiple filters."""
+    filters = {row["filter"]: row["value"] for row in context.table}
+
+    path_pattern = filters.get("path_pattern")
+    file_type_str = filters.get("file_type")
+    tier_str = filters.get("tier")
+
+    file_type = FileType(file_type_str) if file_type_str else None
+    tier = TierLevel(tier_str) if tier_str else None
+
+    context.query_results = context.index.query_combined(
+        path_pattern=path_pattern,
+        file_type=file_type,
+        tier=tier,
+    )

features/steps/actor_cli_yaml_steps.py

@@ -337,6 +337,63 @@ def step_remove_namespaced_ok(context: Any) -> None:
     )
 
 
+@when("I run actor remove with format json")
+def step_remove_format_json(context: Any) -> None:
+    with (
+        patch("cleveragents.cli.commands.actor._get_services") as mock_svc,
+        patch("cleveragents.cli.commands.actor._compute_actor_impact") as mock_impact,
+    ):
+        mock_registry = MagicMock()
+        mock_service = MagicMock()
+        actor = _make_actor(
+            name="local/remove-json",
+            provider="json-provider",
+            model="gpt-json",
+        )
+        mock_registry.get_actor.return_value = actor
+        mock_impact.return_value = (2, 1, 3)
+        mock_svc.return_value = (mock_service, mock_registry)
+
+        context.result = context.runner.invoke(
+            actor_app,
+            ["remove", actor.name, "--format", "json"],
+        )
+
+    context.mock_actor_registry = mock_registry
+    context.actor = actor
+    context.impact_counts = (2, 1, 3)
+
+
+@then("the actor remove output should be valid JSON envelope")
+def step_remove_json_valid(context: Any) -> None:
+    assert context.result.exit_code == 0
+    parsed = json.loads(context.result.output.strip())
+    assert _ENVELOPE_KEYS.issubset(parsed.keys())
+    assert parsed["command"] == f"agents actor remove {context.actor.name}"
+    assert parsed["status"] == "ok"
+    assert parsed["exit_code"] == 0
+    data = _unwrap_envelope(parsed)
+    assert isinstance(data, dict)
+    actor_data = data.get("actor_removed", {})
+    assert actor_data.get("name") == context.actor.name
+    assert actor_data.get("provider") == context.actor.provider
+    assert actor_data.get("model") == context.actor.model
+    impact = data.get("impact", {})
+    expected_sessions, expected_plans, expected_actions = context.impact_counts
+    assert impact.get("sessions") == expected_sessions
+    assert impact.get("active_plans") == expected_plans
+    assert impact.get("actions_referencing") == expected_actions
+    cleanup = data.get("cleanup", {})
+    assert cleanup.get("config") == "kept on disk"
+    assert cleanup.get("contexts") == "0 orphaned"
+    messages = parsed.get("messages", [])
+    assert messages, "expected messages in envelope"
+    first_message = messages[0]
+    assert first_message.get("level") == "ok"
+    assert "Actor removed" in first_message.get("text", "")
+    context.mock_actor_registry.remove_actor.assert_called_once_with(context.actor.name)
+
+
 # ------------------------------------------------------------------
 # Update with --format
 # ------------------------------------------------------------------

features/steps/actor_compiler_coverage_steps.py

@@ -135,7 +135,8 @@ def _subgraph_node(node_id: str, actor_ref: str = "") -> NodeDefinition:
         type=NodeType.SUBGRAPH,
         name=node_id.title(),
         description=f"Subgraph {node_id}",
-        config={"actor_ref": actor_ref},
+        config={},
+        actor_ref=actor_ref if actor_ref else None,
     )
 
 

features/steps/actor_compiler_steps.py

@@ -186,7 +186,8 @@ def step_given_subgraph_ref(context: Context, ref_name: str) -> None:
             type=NodeType.SUBGRAPH,
             name="Sub",
             description="Subgraph",
-            config={"actor_ref": ref_name},
+            config={},
+            actor_ref=ref_name,
         ),
     ]
     edges = [EdgeDefinition(from_node="main", to_node="sub")]
@@ -241,7 +242,8 @@ def step_given_outer_referencing_inner(
             type=NodeType.SUBGRAPH,
             name="Sub",
             description="Subgraph",
-            config={"actor_ref": inner_name},
+            config={},
+            actor_ref=inner_name,
         ),
     ]
     edges = [EdgeDefinition(from_node="main", to_node="sub")]
@@ -260,7 +262,8 @@ def step_given_resolver_cycle(context: Context, inner_name: str, back_ref: str)
             type=NodeType.SUBGRAPH,
             name="Child",
             description="Back-ref",
-            config={"actor_ref": back_ref},
+            config={},
+            actor_ref=back_ref,
         ),
     ]
     inner = _build_graph_config(inner_name, inner_nodes, [], "child", ["child"])

features/steps/actor_subgraph_cycle_detection_steps.py

@@ -0,0 +1,201 @@
+"""Step definitions for cross-actor subgraph cycle detection tests.
+
+Tests for features/actor_subgraph_cycle_detection.feature — validates that
+the actor compiler correctly reads actor_ref from the top-level
+NodeDefinition field (not from node.config) when detecting cross-actor
+subgraph cycles.
+
+This is the regression test for issue #1431: _detect_subgraph_cycles()
+was reading node.config.get("actor_ref", "") instead of node.actor_ref,
+causing cycle detection to always return an empty string and never detect
+cross-actor cycles.
+"""
+
+from __future__ import annotations
+
+from behave import given, then, when
+from behave.runner import Context
+
+from cleveragents.actor.compiler import (
+    SubgraphCycleError,
+    compile_actor,
+)
+from cleveragents.actor.schema import (
+    ActorConfigSchema,
+    ActorType,
+    EdgeDefinition,
+    NodeDefinition,
+    NodeType,
+    RouteDefinition,
+)
+
+
+# ────────────────────────────────────────────────────────────
+# Helpers
+# ────────────────────────────────────────────────────────────
+
+
+def _make_graph_actor(
+    name: str,
+    nodes: list[NodeDefinition],
+    edges: list[EdgeDefinition],
+    entry_node: str,
+    exit_nodes: list[str],
+) -> ActorConfigSchema:
+    """Build a minimal valid GRAPH actor for testing."""
+    route = RouteDefinition(
+        nodes=nodes,
+        edges=edges,
+        entry_node=entry_node,
+        exit_nodes=exit_nodes,
+    )
+    return ActorConfigSchema(
+        name=name,
+        type=ActorType.GRAPH,
+        description=f"Test actor {name}",
+        provider="openai",
+        model="gpt-4",
+        route=route,
+    )
+
+
+def _make_agent_node(node_id: str) -> NodeDefinition:
+    """Build a minimal AGENT node."""
+    return NodeDefinition(
+        id=node_id,
+        type=NodeType.AGENT,
+        name=node_id.title(),
+        description=f"Agent {node_id}",
+        config={"agent": "default"},
+    )
+
+
+def _make_subgraph_node(node_id: str, actor_ref: str) -> NodeDefinition:
+    """Build a SUBGRAPH node using the top-level actor_ref field.
+
+    The actor_ref is set as a first-class field on NodeDefinition, NOT
+    inside config.  This is the correct way to reference a subgraph actor.
+    """
+    return NodeDefinition(
+        id=node_id,
+        type=NodeType.SUBGRAPH,
+        name=node_id.title(),
+        description=f"Subgraph {node_id}",
+        config={},
+        actor_ref=actor_ref,
+    )
+
+
+# ────────────────────────────────────────────────────────────
+# Given steps
+# ────────────────────────────────────────────────────────────
+
+
+@given("the actor compiler is available")
+def step_compiler_available(context: Context) -> None:
+    """Ensure the actor compiler module is importable."""
+    context.actor_registry: dict[str, ActorConfigSchema] = {}
+
+
+@given('actor "{name}" has a subgraph node with actor_ref "{ref}"')
+def step_actor_has_subgraph_ref(context: Context, name: str, ref: str) -> None:
+    """Create a GRAPH actor with one agent node and one subgraph node."""
+    agent = _make_agent_node("start")
+    sub = _make_subgraph_node("sub", actor_ref=ref)
+    actor = _make_graph_actor(
+        name,
+        nodes=[agent, sub],
+        edges=[EdgeDefinition(from_node="start", to_node="sub")],
+        entry_node="start",
+        exit_nodes=["sub"],
+    )
+    context.actor_registry[name] = actor
+    context.actor_to_compile = name
+
+
+@given('actor "{name}" has no subgraph nodes')
+def step_actor_has_no_subgraph(context: Context, name: str) -> None:
+    """Create a GRAPH actor with only an agent node (no subgraph references)."""
+    agent = _make_agent_node("leaf")
+    actor = _make_graph_actor(
+        name,
+        nodes=[agent],
+        edges=[],
+        entry_node="leaf",
+        exit_nodes=["leaf"],
+    )
+    context.actor_registry[name] = actor
+
+
+@given("a registry containing both actors")
+def step_registry_contains_both(context: Context) -> None:
+    """Set up the resolver from the accumulated registry."""
+    registry = context.actor_registry
+
+    def resolver(actor_name: str) -> ActorConfigSchema | None:
+        return registry.get(actor_name)
+
+    context.resolver = resolver
+
+
+# ────────────────────────────────────────────────────────────
+# When steps
+# ────────────────────────────────────────────────────────────
+
+
+@when('I compile actor "{name}" with the registry resolver')
+def step_compile_actor(context: Context, name: str) -> None:
+    """Compile the named actor using the registry resolver."""
+    context.compile_error = None
+    context.compiled = None
+    actor = context.actor_registry[name]
+    try:
+        context.compiled = compile_actor(actor, actor_resolver=context.resolver)
+    except Exception as exc:
+        context.compile_error = exc
+
+
+# ────────────────────────────────────────────────────────────
+# Then steps
+# ────────────────────────────────────────────────────────────
+
+
+@then("the compilation should raise SubgraphCycleError")
+def step_raises_cycle_error(context: Context) -> None:
+    """Assert that compilation raised SubgraphCycleError."""
+    assert context.compile_error is not None, (
+        "Expected SubgraphCycleError but compilation succeeded"
+    )
+    assert isinstance(context.compile_error, SubgraphCycleError), (
+        f"Expected SubgraphCycleError, got {type(context.compile_error).__name__}: "
+        f"{context.compile_error}"
+    )
+
+
+@then("the actor compilation should succeed")
+def step_actor_compilation_succeeds(context: Context) -> None:
+    """Assert that compilation succeeded without errors."""
+    assert context.compile_error is None, (
+        f"Expected compilation to succeed but got: {context.compile_error}"
+    )
+    assert context.compiled is not None
+
+
+@then('the cycle error message should mention "{text}"')
+def step_cycle_error_mentions(context: Context, text: str) -> None:
+    """Assert that the cycle error message contains the given text."""
+    assert context.compile_error is not None
+    msg = str(context.compile_error)
+    assert text.lower() in msg.lower(), (
+        f"Expected '{text}' in error message, got: {msg}"
+    )
+
+
+@then('the actor subgraph_refs should map "{node}" to "{ref}"')
+def step_actor_subgraph_refs_map(context: Context, node: str, ref: str) -> None:
+    """Assert that the compiled metadata maps the node to the expected actor ref."""
+    assert context.compiled is not None
+    refs = context.compiled.metadata.subgraph_refs
+    assert refs.get(node) == ref, (
+        f"Expected subgraph_refs[{node!r}] == {ref!r}, got {refs}"
+    )

features/steps/llm_trace_steps.py

@@ -809,6 +809,9 @@ class _BrokenSession:
     def rollback(self) -> None:
         pass
 
+    def close(self) -> None:
+        pass
+
     def query(self, *_args: Any, **_kwargs: Any) -> Any:
         raise SQLAlchemyDatabaseError("mock", {}, Exception("broken"))
 
@@ -985,8 +988,10 @@ def step_save_with_spy(context: Context) -> None:
     object.__setattr__(real_session, "flush", spy_flush)
     object.__setattr__(real_session, "commit", spy_commit)
 
+    # Pass the session explicitly to test the UoW path: save() must flush
+    # but must NOT commit (the caller owns the transaction boundary).
     repo = LLMTraceRepository(session_factory=lambda: real_session)
-    repo.save(context.trace)
+    repo.save(context.trace, session=real_session)
     # Commit so the data is visible for subsequent queries
     object.__setattr__(real_session, "commit", original_commit)
     real_session.commit()
@@ -1030,7 +1035,9 @@ def step_save_in_uow_rollback(context: Context) -> None:
     session = context.uow_session_factory()
     repo = LLMTraceRepository(session_factory=lambda: session)
     try:
-        repo.save(context.trace)
+        # Pass the session explicitly to use UoW mode: save() flushes but
+        # does NOT commit, so the caller's rollback can undo the change.
+        repo.save(context.trace, session=session)
         # Simulate a subsequent failure that triggers rollback
         raise RuntimeError("Simulated failure after save")
     except RuntimeError:

features/steps/migration_runner_steps.py

@@ -316,6 +316,17 @@ def step_then_temp_connection_closed(context) -> None:
     assert context.current_rev_fake_engine.connections[0].exit_called is True
 
 
+@then("the SQLite engine for get_current_revision should use check_same_thread=False")
+def step_then_get_current_revision_check_same_thread(context) -> None:
+    _url, kwargs = context.current_rev_create_call
+    assert "connect_args" in kwargs, (
+        "Expected connect_args to be passed to create_engine for SQLite"
+    )
+    assert kwargs["connect_args"].get("check_same_thread") is False, (
+        "Expected check_same_thread=False in connect_args for SQLite engine"
+    )
+
+
 @when("I initialize or upgrade a file-based SQLite database")
 def step_when_init_file_based_sqlite(context) -> None:
     import shutil

features/steps/pr_compliance_checklist_steps.py

@@ -0,0 +1,185 @@
+"""Step definitions for PR compliance checklist in implementation supervisor."""
+
+from pathlib import Path
+from typing import Any
+
+from behave import given, then, when
+
+PROJECT_ROOT = Path(__file__).resolve().parents[3]
+AGENT_DEF_PATH = PROJECT_ROOT / ".opencode" / "agents" / "implementation-supervisor.md"
+
+
+@given("the implementation-supervisor.md agent definition exists")
+def step_agent_def_exists(context: Any) -> None:
+    """Verify the implementation supervisor agent definition file exists."""
+    assert AGENT_DEF_PATH.exists(), f"Agent definition not found at {AGENT_DEF_PATH}"
+    context.agent_def_path = AGENT_DEF_PATH
+
+
+@when("I read the implementation supervisor agent definition")
+def step_read_agent_def(context: Any) -> None:
+    """Read the implementation supervisor agent definition."""
+    context.agent_def_content = AGENT_DEF_PATH.read_text(encoding="utf-8")
+
+
+@then("the worker prompt body includes the PR compliance checklist section")
+def step_prompt_includes_checklist(context: Any) -> None:
+    """Verify the worker prompt body includes the PR compliance checklist."""
+    assert "PR Compliance Checklist" in context.agent_def_content, (
+        "Worker prompt body does not include 'PR Compliance Checklist'"
+    )
+
+
+@then("the checklist is marked as MANDATORY")
+def step_checklist_is_mandatory(context: Any) -> None:
+    """Verify the checklist is marked as MANDATORY."""
+    assert "MANDATORY" in context.agent_def_content, (
+        "PR Compliance Checklist is not marked as MANDATORY"
+    )
+
+
+@then("the worker prompt body includes a CHANGELOG.md checklist item")
+def step_prompt_includes_changelog_item(context: Any) -> None:
+    """Verify the worker prompt body includes a CHANGELOG.md checklist item."""
+    assert "CHANGELOG.md" in context.agent_def_content, (
+        "Worker prompt body does not include a CHANGELOG.md checklist item"
+    )
+
+
+@then("the item instructs workers to add an entry under the Unreleased section")
+def step_changelog_item_unreleased(context: Any) -> None:
+    """Verify the CHANGELOG.md item mentions the Unreleased section."""
+    assert "[Unreleased]" in context.agent_def_content, (
+        "CHANGELOG.md checklist item does not mention the [Unreleased] section"
+    )
+
+
+@then("the worker prompt body includes a CONTRIBUTORS.md checklist item")
+def step_prompt_includes_contributors_item(context: Any) -> None:
+    """Verify the worker prompt body includes a CONTRIBUTORS.md checklist item."""
+    assert "CONTRIBUTORS.md" in context.agent_def_content, (
+        "Worker prompt body does not include a CONTRIBUTORS.md checklist item"
+    )
+
+
+@then("the item instructs workers to add or update their contribution entry")
+def step_contributors_item_add_update(context: Any) -> None:
+    """Verify the CONTRIBUTORS.md item instructs workers to add or update."""
+    assert "add or update" in context.agent_def_content, (
+        "CONTRIBUTORS.md checklist item does not instruct workers to add or update"
+    )
+
+
+@then("the worker prompt body includes a commit footer checklist item")
+def step_prompt_includes_commit_footer_item(context: Any) -> None:
+    """Verify the worker prompt body includes a commit footer checklist item."""
+    assert "Commit footer" in context.agent_def_content, (
+        "Worker prompt body does not include a commit footer checklist item"
+    )
+
+
+@then("the item specifies the ISSUES CLOSED footer format")
+def step_commit_footer_issues_closed(context: Any) -> None:
+    """Verify the commit footer item specifies the ISSUES CLOSED format."""
+    assert "ISSUES CLOSED" in context.agent_def_content, (
+        "Commit footer checklist item does not specify the ISSUES CLOSED format"
+    )
+
+
+@then("the worker prompt body includes a CI passes checklist item")
+def step_prompt_includes_ci_item(context: Any) -> None:
+    """Verify the worker prompt body includes a CI passes checklist item."""
+    assert "CI passes" in context.agent_def_content, (
+        "Worker prompt body does not include a CI passes checklist item"
+    )
+
+
+@then("the item instructs workers to verify all quality gates are green")
+def step_ci_item_quality_gates(context: Any) -> None:
+    """Verify the CI item instructs workers to verify quality gates."""
+    assert "quality gates" in context.agent_def_content, (
+        "CI checklist item does not mention quality gates"
+    )
+
+
+@then("the worker prompt body includes a BDD tests checklist item")
+def step_prompt_includes_bdd_item(context: Any) -> None:
+    """Verify the worker prompt body includes a BDD/Behave tests checklist item."""
+    assert "BDD/Behave tests" in context.agent_def_content, (
+        "Worker prompt body does not include a BDD/Behave tests checklist item"
+    )
+
+
+@then("the item instructs workers to add or update Behave feature files")
+def step_bdd_item_feature_files(context: Any) -> None:
+    """Verify the BDD item instructs workers to add or update feature files."""
+    assert "added or updated" in context.agent_def_content, (
+        "BDD checklist item does not instruct workers to add or update feature files"
+    )
+
+
+@then("the worker prompt body includes an Epic reference checklist item")
+def step_prompt_includes_epic_item(context: Any) -> None:
+    """Verify the worker prompt body includes an Epic reference checklist item."""
+    assert "Epic reference" in context.agent_def_content, (
+        "Worker prompt body does not include an Epic reference checklist item"
+    )
+
+
+@then("the item instructs workers to reference the parent Epic issue number")
+def step_epic_item_parent_reference(context: Any) -> None:
+    """Verify the Epic item instructs workers to reference the parent Epic."""
+    assert "parent Epic" in context.agent_def_content, (
+        "Epic checklist item does not instruct workers to reference the parent Epic"
+    )
+
+
+@then("the worker prompt body includes a labels checklist item")
+def step_prompt_includes_labels_item(context: Any) -> None:
+    """Verify the worker prompt body includes a labels checklist item."""
+    assert "Labels" in context.agent_def_content, (
+        "Worker prompt body does not include a labels checklist item"
+    )
+
+
+@then("the item instructs workers to apply labels via forgejo-label-manager")
+def step_labels_item_forgejo_label_manager(context: Any) -> None:
+    """Verify the labels item instructs workers to use forgejo-label-manager."""
+    assert "forgejo-label-manager" in context.agent_def_content, (
+        "Labels checklist item does not mention forgejo-label-manager"
+    )
+
+
+@then("the worker prompt body includes a milestone checklist item")
+def step_prompt_includes_milestone_item(context: Any) -> None:
+    """Verify the worker prompt body includes a milestone checklist item."""
+    assert "Milestone" in context.agent_def_content, (
+        "Worker prompt body does not include a milestone checklist item"
+    )
+
+
+@then("the item instructs workers to assign the earliest open milestone")
+def step_milestone_item_earliest(context: Any) -> None:
+    """Verify the milestone item instructs workers to assign the earliest open milestone."""
+    assert "earliest open milestone" in context.agent_def_content, (
+        "Milestone checklist item does not mention the earliest open milestone"
+    )
+
+
+@then("the worker prompt body contains all 8 mandatory checklist items")
+def step_prompt_contains_all_8_items(context: Any) -> None:
+    """Verify the worker prompt body contains all 8 mandatory checklist items."""
+    required_items = [
+        "CHANGELOG.md",
+        "CONTRIBUTORS.md",
+        "ISSUES CLOSED",
+        "CI passes",
+        "BDD/Behave tests",
+        "Epic reference",
+        "forgejo-label-manager",
+        "earliest open milestone",
+    ]
+    missing = [item for item in required_items if item not in context.agent_def_content]
+    assert not missing, (
+        f"Worker prompt body is missing the following checklist items: {missing}"
+    )

features/steps/session_cli_steps.py

@@ -4,6 +4,7 @@ from __future__ import annotations
 
 import json
 import os
+import re
 import tempfile
 from datetime import datetime
 from typing import Any
@@ -160,14 +161,34 @@ def step_existing_sessions(context: Context) -> None:
     context.mock_service.list.return_value = sessions
 
 
+@given("there are mocked existing named sessions")
+def step_existing_named_sessions(context: Context) -> None:
+    """Set up mocked sessions with names for Summary panel name-display test."""
+    sessions = [
+        _make_session(
+            session_id=_SESSION_ID,
+            actor_name="openai/gpt-4",
+            messages=[_make_message(sequence=0)],
+        ),
+        _make_session(session_id=_SESSION_ID_2),
+    ]
+    sessions[0].name = "weekly-planning"
+    sessions[1].name = "refactor-sprint"
+    context.mock_service.list.return_value = sessions
+
+
 @when("I run session CLI list")
 def step_list(context: Context) -> None:
-    context.result = context.runner.invoke(session_app, ["list"])
+    context.result = context.runner.invoke(
+        session_app, ["list"], env={"COLUMNS": "200"}
+    )
 
 
 @when("I run session CLI list with --format json")
 def step_list_json(context: Context) -> None:
-    context.result = context.runner.invoke(session_app, ["list", "--format", "json"])
+    context.result = context.runner.invoke(
+        session_app, ["list", "--format", "json"], env={"COLUMNS": "200"}
+    )
 
 
 @then("the session CLI should show all sessions in a table")
@@ -176,6 +197,125 @@ def step_list_shows_table(context: Context) -> None:
     assert "Sessions" in context.result.output
 
 
+@then("the session CLI rich table should display full session ULIDs")
+def step_list_rich_table_full_ulids(context: Context) -> None:
+    """Verify the Rich table displays the full 26-character session ULIDs."""
+    assert context.result.exit_code == 0
+    output = context.result.output
+    # Restrict assertion to the table region (before the Summary panel) so
+    # that a regression back to 8-char truncation is not masked by the full
+    # ULID appearing elsewhere (e.g. in the Summary panel).
+    summary_idx = output.find("Summary")
+    table_output = output[:summary_idx] if summary_idx != -1 else output
+    assert _SESSION_ID in table_output, (
+        f"Full ULID {_SESSION_ID} not found in Rich table output:\n"
+        f"{context.result.output}"
+    )
+    assert _SESSION_ID_2 in table_output, (
+        f"Full ULID {_SESSION_ID_2} not found in Rich table output:\n"
+        f"{context.result.output}"
+    )
+    # Negative guard: ensure the table contains full 26-character ULIDs,
+    # not the old 8-character truncated form.  A standalone 8-char prefix
+    # (not as part of the full ULID) would indicate the [:8] slice was not
+    # removed.
+    table_without_full_ids = table_output.replace(_SESSION_ID, "").replace(
+        _SESSION_ID_2, ""
+    )
+    assert _SESSION_ID[:8] not in table_without_full_ids, (
+        f"Truncated 8-char ID {_SESSION_ID[:8]} found in Rich table output:\n"
+        f"{context.result.output}"
+    )
+    assert _SESSION_ID_2[:8] not in table_without_full_ids, (
+        f"Truncated 8-char ID {_SESSION_ID_2[:8]} found in Rich table output:\n"
+        f"{context.result.output}"
+    )
+
+
+@then("the session CLI summary panel should contain full session ULIDs")
+def step_list_summary_full_ulids(context: Context) -> None:
+    """Verify the Summary panel shows full ULIDs for unnamed sessions."""
+    assert context.result.exit_code == 0
+    output = context.result.output
+    assert "Summary" in output, f"Summary panel not found in output:\n{output}"
+    # Extract the Summary panel region to avoid a false pass from the Rich
+    # table also containing the same full ULIDs.
+    summary_idx = output.find("Summary")
+    summary_output = output[summary_idx:]
+    # Both Most Recent and Oldest entries should display full ULIDs when
+    # sessions are unnamed.  Asserting only one ID would allow a regression
+    # that re-introduced [:8] on one fallback path while keeping the other
+    # intact to pass the test undetected.
+    assert _SESSION_ID in summary_output, (
+        f"Full ULID {_SESSION_ID} not found in Summary panel:\n{output}"
+    )
+    assert _SESSION_ID_2 in summary_output, (
+        f"Full ULID {_SESSION_ID_2} not found in Summary panel:\n{output}"
+    )
+
+
+@then("the session CLI summary panel should show session names")
+def step_list_summary_shows_names(context: Context) -> None:
+    """Verify the Summary panel shows session names (not ULIDs) for named sessions."""
+    assert context.result.exit_code == 0
+    output = context.result.output
+    summary_idx = output.find("Summary")
+    assert summary_idx != -1, f"Summary panel not found in output:\n{output}"
+    summary_output = output[summary_idx:]
+    assert "weekly-planning" in summary_output, (
+        f"'weekly-planning' not found in Summary panel:\n{output}"
+    )
+    assert "refactor-sprint" in summary_output, (
+        f"'refactor-sprint' not found in Summary panel:\n{output}"
+    )
+    # The Summary should NOT show the raw ULID when session names are present
+    assert _SESSION_ID not in summary_output, (
+        f"Full ULID {_SESSION_ID} unexpectedly found in Summary panel:\n{output}"
+    )
+    assert _SESSION_ID_2 not in summary_output, (
+        f"Full ULID {_SESSION_ID_2} unexpectedly found in Summary panel:\n{output}"
+    )
+
+
+@when("I capture the first session full ULID from the output")
+def step_capture_first_ulid(context: Context) -> None:
+    """Parse the first session's full ULID from the output and store it
+    for subsequent steps (round-trip tell test)."""
+    assert context.result.exit_code == 0
+    output = context.result.output
+    # Restrict the search to the table region (before the Summary panel)
+    # so that a ULID appearing in the Summary panel is not accidentally
+    # captured as the "first" session ID.
+    summary_idx = output.find("Summary")
+    search_region = output[:summary_idx] if summary_idx != -1 else output
+    match = re.search(r"[0-9A-HJKMNP-TV-Z]{26}", search_region)
+    assert match is not None, (
+        f"No 26-character ULID found in table output:\n{search_region}"
+    )
+    context.full_session_id = match.group()
+    assert len(context.full_session_id) == 26, (
+        f"Parsed ULID '{context.full_session_id}' is not 26 characters"
+    )
+    # Sanity check: the captured ID should match one of the known fixture IDs
+    assert context.full_session_id in (_SESSION_ID, _SESSION_ID_2), (
+        f"Captured ULID '{context.full_session_id}' does not match any fixture ID"
+    )
+
+
+@when('I run session CLI tell with the full session ID and prompt "{prompt}"')
+def step_tell_with_stored_ulid(context: Context, prompt: str) -> None:
+    """Run session tell using the full ULID stored from session list."""
+    session_id = context.full_session_id
+    context.mock_service.append_message.side_effect = [
+        _make_message(MessageRole.USER, prompt, 0),
+        _make_message(MessageRole.ASSISTANT, f"Acknowledged: {prompt}", 1),
+    ]
+    context.result = context.runner.invoke(
+        session_app,
+        ["tell", "--session", session_id, prompt],
+    )
+
+
 # ---------------------------------------------------------------------------
 # Show
 # ---------------------------------------------------------------------------

features/steps/tdd_migration_runner_get_current_revision_threading_10507_steps.py

@@ -0,0 +1,85 @@
+"""Steps for TDD Issue #10507 — get_current_revision() SQLite threading fix."""
+
+from __future__ import annotations
+
+from typing import Any
+from unittest.mock import MagicMock, patch
+
+from behave import then, when
+
+
+def _run_get_current_revision_and_capture_kwargs(
+    context: Any,
+) -> None:
+    """Shared helper: call get_current_revision() and capture create_engine kwargs.
+
+    Mocks ``create_engine`` and ``MigrationContext.configure`` so the call
+    completes without a real database.  The keyword arguments passed to
+    ``create_engine`` are stored on ``context.engine_creation_kwargs`` for
+    subsequent assertion steps.
+    """
+    fake_engine = MagicMock()
+    fake_connection = MagicMock()
+    fake_connection.__enter__ = MagicMock(return_value=fake_connection)
+    fake_connection.__exit__ = MagicMock(return_value=False)
+    fake_engine.connect.return_value = fake_connection
+
+    migration_ctx = MagicMock()
+    migration_ctx.get_current_revision.return_value = None
+
+    captured_kwargs: list[dict[str, Any]] = []
+
+    def fake_create_engine(url: str, **kwargs: Any) -> MagicMock:
+        captured_kwargs.append(kwargs)
+        return fake_engine
+
+    with (
+        patch(
+            "cleveragents.infrastructure.database.migration_runner.create_engine",
+            side_effect=fake_create_engine,
+        ),
+        patch(
+            "cleveragents.infrastructure.database.migration_runner.MigrationContext.configure",
+            return_value=migration_ctx,
+        ),
+    ):
+        context.revision_result = context.runner.get_current_revision()
+
+    context.engine_creation_kwargs = captured_kwargs
+
+
+@when("I request the current revision and capture the engine creation args")
+def step_when_capture_engine_args(context: Any) -> None:
+    """Call get_current_revision and capture the create_engine call arguments."""
+    _run_get_current_revision_and_capture_kwargs(context)
+
+
+@then("the SQLite engine should be created with check_same_thread set to False")
+def step_then_sqlite_engine_has_check_same_thread(context: Any) -> None:
+    """Verify the SQLite engine was created with check_same_thread=False."""
+    assert len(context.engine_creation_kwargs) == 1, (
+        f"Expected exactly 1 create_engine call, got {len(context.engine_creation_kwargs)}"
+    )
+    kwargs = context.engine_creation_kwargs[0]
+    assert "connect_args" in kwargs, (
+        "Expected connect_args in create_engine kwargs for SQLite, "
+        f"but got kwargs: {kwargs}"
+    )
+    assert kwargs["connect_args"].get("check_same_thread") is False, (
+        "Expected check_same_thread=False in connect_args, "
+        f"but got: {kwargs['connect_args']}"
+    )
+
+
+@then("the non-SQLite engine should be created without check_same_thread")
+def step_then_non_sqlite_engine_no_check_same_thread(context: Any) -> None:
+    """Verify non-SQLite engines are not given check_same_thread."""
+    assert len(context.engine_creation_kwargs) == 1, (
+        f"Expected exactly 1 create_engine call, got {len(context.engine_creation_kwargs)}"
+    )
+    kwargs = context.engine_creation_kwargs[0]
+    connect_args = kwargs.get("connect_args", {})
+    assert "check_same_thread" not in connect_args, (
+        "Expected check_same_thread to be absent for non-SQLite engine, "
+        f"but got connect_args: {connect_args}"
+    )

features/steps/tui_session_persistence_tabs_steps.py

@@ -0,0 +1,293 @@
+"""Step definitions for TUI session persistence and tab bar tests."""
+
+from __future__ import annotations
+
+import tempfile
+from pathlib import Path
+from typing import Any
+
+from behave import given, then, use_step_matcher, when
+
+from cleveragents.tui.session_store import SessionStore
+from cleveragents.tui.widgets.session_tab_bar import SessionTabBar
+
+
+def _get_widget_text(widget: SessionTabBar) -> str:
+    """Safely extract text from a TabBar widget.
+
+    Handles the Textual Static variant (via internal _current_text), and
+    the fallback implementation (_text attribute).
+
+    Args:
+        widget: The tab bar widget instance.
+
+    Returns:
+        The rendered text content as a string, or empty string on failure.
+    """
+    if widget is None:
+        return ""
+
+    # SessionTabBar tracks its own internal text for Textual Static case.
+    current_text = getattr(widget, "_current_text", None)
+    if current_text is not None:
+        return str(current_text)
+
+    # Fall back to the fallback implementation's _text.
+    text_attr = getattr(widget, "_text", None)
+    if text_attr is not None:
+        return str(text_attr)
+
+    return ""
+
+
+@given("a clean TUI session store")
+def step_clean_session_store(context: Any) -> None:
+    """Create a clean session store using an in-memory temp database."""
+    temp_dir = tempfile.mkdtemp()
+    db_path = Path(temp_dir) / "tui.db"
+    context.session_store = SessionStore(db_path=db_path)
+    context.sessions: list[dict[str, str]] = []
+    context.tab_bar: SessionTabBar | None = None
+
+
+@given("a clean TUI tab bar")
+def step_clean_tab_bar(context: Any) -> None:
+    """Create a clean session tab bar widget."""
+    context.tab_bar = SessionTabBar()
+
+
+# ---------------------------------------------------------------------------
+# Session CRUD steps
+# ---------------------------------------------------------------------------
+
+# Use regex matcher to prevent AmbiguousStep: {name} in parse mode is greedy
+# enough to consume the trailing ' with state "..."' suffix, making the two
+# patterns overlap. Explicit [^"]+ anchors the capture to within the quotes.
+use_step_matcher("re")
+
+
+@when(
+    r'I create a session with id "(?P<session_id>[^"]+)"'
+    r' and name "(?P<name>[^"]+)"'
+)
+def step_create_session(context: Any, session_id: str, name: str) -> None:
+    """Create a session in the store."""
+    if context.session_store is None:  # type: ignore[possibly-undefined]
+        raise RuntimeError("session_store not initialised")
+    record = context.session_store.create_session(session_id, name)
+    context.sessions.append(
+        {
+            "session_id": record.session_id,
+            "name": record.name,
+            "state": record.state,
+        }
+    )
+
+
+@when(
+    r'I create a session with id "(?P<session_id>[^"]+)"'
+    r' and name "(?P<name>[^"]+)"'
+    r' with state "(?P<state>[^"]+)"'
+)
+def step_create_session_with_state(
+    context: Any, session_id: str, name: str, state: str
+) -> None:
+    """Create a session with a specific initial state."""
+    if context.session_store is None:  # type: ignore[possibly-undefined]
+        raise RuntimeError("session_store not initialised")
+    record = context.session_store.create_session(session_id, name, state=state)
+    context.sessions.append(
+        {
+            "session_id": record.session_id,
+            "name": record.name,
+            "state": record.state,
+        }
+    )
+
+
+use_step_matcher("parse")
+
+
+@then("the session should be persisted in the database")
+def step_session_persisted(context: Any) -> None:
+    """Verify that at least one session was created."""
+    assert len(context.sessions) > 0, "No sessions were created"
+    last = context.sessions[-1]
+    retrieved = context.session_store.get_session(last["session_id"])
+    assert retrieved is not None, f"Session {last['session_id']} not found in DB"
+
+
+@then("I should be able to retrieve the session by id")
+def step_retrieve_session(context: Any) -> None:
+    """Verify retrieval returns matching data."""
+    last = context.sessions[-1]
+    retrieved = context.session_store.get_session(last["session_id"])
+    assert retrieved is not None
+    assert retrieved.session_id == last["session_id"]
+    assert retrieved.name == last["name"]
+
+
+@then("I should have {count:d} sessions in the store")
+def step_count_sessions(context: Any, count: int) -> None:
+    """Verify the number of sessions persisted."""
+    all_sessions = context.session_store.list_sessions()
+    assert len(all_sessions) == count, (
+        f"Expected {count} sessions, got {len(all_sessions)}"
+    )
+
+
+@then("the sessions should be ordered by creation time")
+def step_sessions_ordered(context: Any) -> None:
+    """Verify chronological ordering of persisted sessions."""
+    all_sessions = context.session_store.list_sessions()
+    for i in range(len(all_sessions) - 1):
+        assert all_sessions[i].created_at <= (all_sessions[i + 1].created_at)
+
+
+@when('I update the session state to "{state}"')
+def step_update_session_state(context: Any, state: str) -> None:
+    """Update the last-created session's state."""
+    if context.sessions is None or not context.sessions:
+        raise RuntimeError("No sessions to update")
+    last = context.sessions[-1]
+    context.session_store.update_session_state(last["session_id"], state)
+    last["state"] = state
+
+
+@then('the session state should be "{state}"')
+def step_verify_session_state(context: Any, state: str) -> None:
+    """Verify the persisted state matches expectation."""
+    if context.sessions is None or not context.sessions:
+        raise RuntimeError("No sessions to verify")
+    last = context.sessions[-1]
+    retrieved = context.session_store.get_session(last["session_id"])
+    assert retrieved is not None
+    assert retrieved.state == state
+
+
+@when("I delete the session")
+def step_delete_session(context: Any) -> None:
+    """Delete the last-created session from the store."""
+    if context.sessions is None or not context.sessions:
+        raise RuntimeError("No sessions to delete")
+    last = context.sessions[-1]
+    context.session_store.delete_session(last["session_id"])
+
+
+@then("the session should not exist in the store")
+def step_session_not_exist(context: Any) -> None:
+    """Verify deletion removed the session from storage."""
+    if context.sessions is None or not context.sessions:
+        raise RuntimeError("No sessions to verify deletion")
+    last = context.sessions[-1]
+    retrieved = context.session_store.get_session(last["session_id"])
+    assert retrieved is None
+
+
+# ---------------------------------------------------------------------------
+# Tab bar steps
+# ---------------------------------------------------------------------------
+
+
+@when("I render the tab bar with {count:d} session")
+def step_render_tab_bar_single(context: Any, count: int) -> None:
+    """Render the tab bar with exactly one session (should be hidden)."""
+    if context.tab_bar is None:
+        context.tab_bar = SessionTabBar()
+    context.tab_bar.set_sessions(context.sessions[:count])
+
+
+@when("I render the tab bar with {count:d} sessions")
+def step_render_tab_bar_multiple(context: Any, count: int) -> None:
+    """Render the tab bar with multiple sessions."""
+    if context.tab_bar is None:
+        context.tab_bar = SessionTabBar()
+    context.tab_bar.set_sessions(context.sessions[:count])
+
+
+@when("I render the tab bar with these sessions")
+def step_render_tab_bar_with_sessions(context: Any) -> None:
+    """Render the tab bar using all stored sessions."""
+    if context.tab_bar is None:
+        context.tab_bar = SessionTabBar()
+    context.tab_bar.set_sessions(context.sessions)
+
+
+@when('I render the tab bar with active session "{session_id}"')
+def step_render_tab_bar_with_active(context: Any, session_id: str) -> None:
+    """Render the tab bar highlighting a specific session."""
+    if context.tab_bar is None:
+        context.tab_bar = SessionTabBar()
+    context.tab_bar.set_sessions(context.sessions, active_session_id=session_id)
+
+
+@then("the tab bar should be hidden")
+def step_tab_bar_hidden(context: Any) -> None:
+    """Verify the tab bar widget reports as not visible."""
+    tb = context.tab_bar
+    if tb is None:
+        raise AssertionError("Tab bar was never created")
+    # The real Textual Static uses `.visible`; the fallback uses `.display`.
+    for attr_name in ("visible", "display"):
+        val = getattr(tb, attr_name, None)
+        if val is not None:
+            assert not val, f"Tab bar should be hidden (via {attr_name})"
+            return
+    raise AssertionError("Tab bar has neither visible nor display")
+
+
+@then("the tab bar should be visible")
+def step_tab_bar_visible(context: Any) -> None:
+    """Verify the tab bar widget reports as visible."""
+    tb = context.tab_bar
+    if tb is None:
+        raise AssertionError("Tab bar was never created")
+    for attr_name in ("visible", "display"):
+        val = getattr(tb, attr_name, None)
+        if val is not None:
+            assert val, f"Tab bar should be visible (via {attr_name})"
+            return
+    raise AssertionError("Tab bar has neither visible nor display")
+
+
+@then('the tab bar should show "\\u231b" for the working session')
+def step_tab_bar_working_indicator(context: Any) -> None:
+    """Verify the hourglass indicator appears on a working session."""
+    widget = context.tab_bar
+    if widget is None:
+        raise AssertionError("Tab bar was never created")
+    content = _get_widget_text(widget)
+    assert "\u231b" in content or "⌛" in content, (
+        f"Hourglass indicator not found. Content: {content!r}"
+    )
+
+
+@then('the tab bar should show "\\u276f" for the awaiting_input session')
+def step_tab_bar_awaiting_indicator(context: Any) -> None:
+    """Verify the prompt-arrow indicator appears on an awaiting-input session."""
+    widget = context.tab_bar
+    if widget is None:
+        raise AssertionError("Tab bar was never created")
+    content = _get_widget_text(widget)
+    assert "\u276f" in content, f"Prompt-arrow not found. Content: {content!r}"
+
+
+@then("the tab bar should show no indicator for the idle session")
+def step_tab_bar_idle_no_indicator(context: Any) -> None:
+    """Verify idle sessions have no prefix indicator."""
+    # This is implicitly verified by the presence of indicators on
+    # non-idle sessions in multi-session scenarios.
+    pass
+
+
+@then('the tab bar should mark "{name}" as active with brackets')
+def step_tab_bar_active_marked(context: Any, name: str) -> None:
+    """Verify the designated session is wrapped in square brackets."""
+    widget = context.tab_bar
+    if widget is None:
+        raise AssertionError("Tab bar was never created")
+    content = _get_widget_text(widget)
+    pattern = f"[{name}]"
+    assert pattern in content, (
+        f"Active marker '{pattern}' not found. Content: {content!r}"
+    )

features/tdd_migration_runner_get_current_revision_threading_10507.feature

@@ -0,0 +1,31 @@
+@tdd_issue @tdd_issue_10507
+Feature: TDD Issue #10507 — get_current_revision() must pass check_same_thread=False for SQLite
+  As a developer using MigrationRunner in a multi-threaded application
+  I want get_current_revision() to work safely from background threads
+  So that async startup flows and background migration checks do not crash
+
+  The root cause is that MigrationRunner.get_current_revision() calls
+  create_engine(self.database_url) without connect_args={"check_same_thread": False}
+  for SQLite databases.  When called from a thread other than the one that
+  created the engine, SQLite raises:
+      ProgrammingError: SQLite objects created in a thread can only be
+      used in that same thread.
+
+  The sibling method init_or_upgrade() already passes check_same_thread=False
+  for SQLite, making this an inconsistency in the same class.  Because
+  get_pending_migrations() and check_migrations_needed() both delegate to
+  get_current_revision(), the threading bug propagates to all three methods.
+
+  The fix adds connect_args={"check_same_thread": False} to the create_engine()
+  call in get_current_revision() when the database URL starts with "sqlite",
+  consistent with the existing pattern in init_or_upgrade().
+
+  Scenario: get_current_revision passes check_same_thread=False for SQLite engine
+    Given a migration runner configured for "sqlite:///:memory:"
+    When I request the current revision and capture the engine creation args
+    Then the SQLite engine should be created with check_same_thread set to False
+
+  Scenario: get_current_revision does not pass check_same_thread for non-SQLite engine
+    Given a migration runner configured for "postgresql://user:pass@localhost/testdb"
+    When I request the current revision and capture the engine creation args
+    Then the non-SQLite engine should be created without check_same_thread

features/tui_session_persistence_tabs.feature

@@ -0,0 +1,54 @@
+Feature: TUI Session Persistence and Multi-Session Tab Bar
+  As a user of the CleverAgents TUI
+  I want sessions to persist across restarts
+  And I want to manage multiple sessions with a tab bar
+
+  Background:
+    Given a clean TUI session store
+
+  Scenario: Create and persist a session
+    When I create a session with id "session-1" and name "Main Session"
+    Then the session should be persisted in the database
+    And I should be able to retrieve the session by id
+
+  Scenario: List all sessions
+    When I create a session with id "session-1" and name "Session 1"
+    And I create a session with id "session-2" and name "Session 2"
+    Then I should have 2 sessions in the store
+    And the sessions should be ordered by creation time
+
+  Scenario: Update session state
+    When I create a session with id "session-1" and name "Main Session"
+    And I update the session state to "working"
+    Then the session state should be "working"
+
+  Scenario: Delete a session
+    When I create a session with id "session-1" and name "Main Session"
+    And I delete the session
+    Then the session should not exist in the store
+
+  Scenario: Tab bar is hidden with single session
+    When I create a session with id "session-1" and name "Main Session"
+    And I render the tab bar with 1 session
+    Then the tab bar should be hidden
+
+  Scenario: Tab bar is visible with multiple sessions
+    When I create a session with id "session-1" and name "Session 1"
+    And I create a session with id "session-2" and name "Session 2"
+    And I render the tab bar with 2 sessions
+    Then the tab bar should be visible
+
+  Scenario: Tab bar shows state indicators
+    When I create a session with id "session-1" and name "Session 1" with state "idle"
+    And I create a session with id "session-2" and name "Session 2" with state "working"
+    And I create a session with id "session-3" and name "Session 3" with state "awaiting_input"
+    And I render the tab bar with these sessions
+    Then the tab bar should show "\u231b" for the working session
+    And the tab bar should show "\u276f" for the awaiting_input session
+    And the tab bar should show no indicator for the idle session
+
+  Scenario: Tab bar marks active session
+    When I create a session with id "session-1" and name "Session 1"
+    And I create a session with id "session-2" and name "Session 2"
+    And I render the tab bar with active session "session-1"
+    Then the tab bar should mark "Session 1" as active with brackets

robot/actor_compiler.robot

@@ -42,3 +42,15 @@ Reject LLM Actor Compilation
     Should Be Equal As Integers    ${result.rc}    0
     Should Contain    ${result.stdout}    actor-compiler-expected-fail
     Should Contain    ${result.stdout}    GRAPH
+
+Detect Cross-Actor Subgraph Cycle Via actor_ref Field
+    [Documentation]    Verify that mutually-referencing actors raise SubgraphCycleError.
+    ...    This is the regression test for issue #1431: _detect_subgraph_cycles()
+    ...    was reading node.config.get("actor_ref") instead of node.actor_ref,
+    ...    causing cycle detection to silently fail.
+    [Tags]    slow
+    ${result}=    Run Process    ${PYTHON}    ${HELPER}    cycle-detect    dummy    cwd=${WORKSPACE}
+    Log    ${result.stdout}
+    Log    ${result.stderr}
+    Should Be Equal As Integers    ${result.rc}    0
+    Should Contain    ${result.stdout}    actor-compiler-cycle-detected

robot/actor_remove_cli.robot

@@ -0,0 +1,18 @@
+*** Settings ***
+Documentation    Integration tests for actor remove CLI format output
+Resource         ${CURDIR}/common.resource
+Suite Setup      Setup Test Environment
+Suite Teardown   Cleanup Test Environment
+
+*** Variables ***
+${HELPER}    ${CURDIR}/helper_actor_remove_cli.py
+
+*** Test Cases ***
+Actor Remove Format JSON Emits Spec Envelope
+    [Documentation]    Verify that ``actor remove --format json`` emits a spec-compliant envelope
+    [Tags]    actor_remove_cli_format
+    ${result}=    Run Process    ${PYTHON}    ${HELPER}    remove-json    cwd=${WORKSPACE}
+    Log    ${result.stdout}
+    Log    ${result.stderr}
+    Should Be Equal As Integers    ${result.rc}    0
+    Should Contain    ${result.stdout}    actor-remove-json-format-ok

robot/helper_actor_compiler.py

@@ -66,6 +66,82 @@ def main() -> int:
             print(f"actor-compiler-fail: {exc}")
             return 1
 
+    if command == "cycle-detect":
+        # Test cross-actor cycle detection using actor_ref top-level field.
+        # Creates two mutually-referencing actors and verifies SubgraphCycleError
+        # is raised, confirming the fix for issue #1431.
+        from cleveragents.actor.compiler import SubgraphCycleError
+        from cleveragents.actor.schema import (
+            ActorType,
+            EdgeDefinition,
+            NodeDefinition,
+            NodeType,
+            RouteDefinition,
+        )
+
+        def _make_actor(
+            name: str, subgraph_ref: str | None = None
+        ) -> ActorConfigSchema:
+            if subgraph_ref is not None:
+                nodes = [
+                    NodeDefinition(
+                        id="start",
+                        type=NodeType.AGENT,
+                        name="Start",
+                        description="Start",
+                        config={"agent": "default"},
+                    ),
+                    NodeDefinition(
+                        id="sub",
+                        type=NodeType.SUBGRAPH,
+                        name="Sub",
+                        description="Subgraph",
+                        config={},
+                        actor_ref=subgraph_ref,
+                    ),
+                ]
+                edges = [EdgeDefinition(from_node="start", to_node="sub")]
+                entry, exits = "start", ["sub"]
+            else:
+                nodes = [
+                    NodeDefinition(
+                        id="leaf",
+                        type=NodeType.AGENT,
+                        name="Leaf",
+                        description="Leaf",
+                        config={"agent": "default"},
+                    )
+                ]
+                edges = []
+                entry, exits = "leaf", ["leaf"]
+            route = RouteDefinition(
+                nodes=nodes, edges=edges, entry_node=entry, exit_nodes=exits
+            )
+            return ActorConfigSchema(
+                name=name,
+                type=ActorType.GRAPH,
+                description=f"Test actor {name}",
+                model="gpt-4",
+                route=route,
+            )
+
+        actor_a = _make_actor("test/actor-a", subgraph_ref="test/actor-b")
+        actor_b = _make_actor("test/actor-b", subgraph_ref="test/actor-a")
+        registry = {"test/actor-a": actor_a, "test/actor-b": actor_b}
+
+        try:
+            compile_actor(actor_a, actor_resolver=registry.get)
+            print(
+                "actor-compiler-cycle-not-detected: FAIL — expected SubgraphCycleError"
+            )
+            return 1
+        except SubgraphCycleError as exc:
+            print(f"actor-compiler-cycle-detected: {exc}")
+            return 0
+        except Exception as exc:
+            print(f"actor-compiler-unexpected-error: {exc}")
+            return 1
+
     print(f"Unknown command: {command}")
     return 1
 

robot/helper_actor_remove_cli.py

@@ -0,0 +1,164 @@
+"""Helper script for Robot integration tests covering ``actor remove --format`` output.
+
+Exercises the real ``agents`` CLI via subprocess — no mocking of any kind.
+A test actor is seeded via ``agents actor add``, then removed via
+``agents actor remove --format json``, and the resulting JSON envelope is
+validated against the spec.
+
+Usage::
+
+    python helper_actor_remove_cli.py <command>
+
+Where *command* is one of:
+
+- ``remove-json`` — seed an actor, remove it with ``--format json``, validate envelope
+"""
+
+from __future__ import annotations
+
+import json
+import os
+import sys
+from pathlib import Path
+
+# Ensure src is importable when run from workspace root
+_SRC = str(Path(__file__).resolve().parents[1] / "src")
+if _SRC not in sys.path:
+    sys.path.insert(0, _SRC)
+
+# Ensure robot/ is on the import path for helper_e2e_common.
+_ROBOT = str(Path(__file__).resolve().parent)
+if _ROBOT not in sys.path:
+    sys.path.insert(0, _ROBOT)
+
+from helper_e2e_common import cleanup_workspace, run_cli, setup_workspace  # noqa: E402
+
+_ACTOR_NAME = "local/robot-remove-actor"
+
+_ACTOR_CONFIG: dict[str, object] = {
+    "name": _ACTOR_NAME,
+    "provider": "openai",
+    "model": "gpt-4",
+}
+
+
+def _write_actor_config(workspace: str) -> str:
+    """Write actor config JSON to a temp file and return its path."""
+    config_path = os.path.join(workspace, "robot_remove_actor.json")
+    with open(config_path, "w", encoding="utf-8") as fh:
+        json.dump(_ACTOR_CONFIG, fh)
+    return config_path
+
+
+def test_remove_format_json() -> None:
+    """Seed an actor via the real CLI, remove it with ``--format json``.
+
+    Validates the resulting JSON envelope against the spec.
+    """
+    workspace = setup_workspace(prefix="robot_actor_remove_")
+    try:
+        config_path = _write_actor_config(workspace)
+
+        # Step 1: Add the actor using the real CLI.
+        add_result = run_cli(
+            "actor",
+            "add",
+            _ACTOR_NAME,
+            "--config",
+            config_path,
+            workspace=workspace,
+        )
+        assert add_result.returncode == 0, (
+            f"actor add failed (rc={add_result.returncode}):\n"
+            f"stdout: {add_result.stdout}\nstderr: {add_result.stderr}"
+        )
+
+        # Step 2: Remove the actor with --format json using the real CLI.
+        remove_result = run_cli(
+            "actor",
+            "remove",
+            _ACTOR_NAME,
+            "--format",
+            "json",
+            workspace=workspace,
+        )
+        assert remove_result.returncode == 0, (
+            f"actor remove --format json failed (rc={remove_result.returncode}):\n"
+            f"stdout: {remove_result.stdout}\nstderr: {remove_result.stderr}"
+        )
+
+        # Step 3: Parse and validate the JSON envelope.
+        output = remove_result.stdout.strip()
+        assert output, (
+            f"actor remove --format json produced no output.\n"
+            f"stderr: {remove_result.stderr}"
+        )
+
+        payload = json.loads(output)
+
+        assert payload["command"] == f"agents actor remove {_ACTOR_NAME}", (
+            f"Unexpected command field: {payload.get('command')!r}"
+        )
+        assert payload["status"] == "ok", (
+            f"Unexpected status: {payload.get('status')!r}"
+        )
+        assert payload["exit_code"] == 0, (
+            f"Unexpected exit_code: {payload.get('exit_code')!r}"
+        )
+
+        data = payload["data"]
+
+        removed = data.get("actor_removed", {})
+        assert removed.get("name") == _ACTOR_NAME, (
+            f"actor_removed.name mismatch: {removed.get('name')!r}"
+        )
+        assert removed.get("provider") == _ACTOR_CONFIG["provider"], (
+            f"actor_removed.provider mismatch: {removed.get('provider')!r}"
+        )
+        assert removed.get("model") == _ACTOR_CONFIG["model"], (
+            f"actor_removed.model mismatch: {removed.get('model')!r}"
+        )
+
+        impact = data.get("impact", {})
+        assert "sessions" in impact, f"Missing 'sessions' in impact: {impact}"
+        assert "active_plans" in impact, f"Missing 'active_plans' in impact: {impact}"
+        assert "actions_referencing" in impact, (
+            f"Missing 'actions_referencing' in impact: {impact}"
+        )
+
+        cleanup = data.get("cleanup", {})
+        assert cleanup.get("config") == "kept on disk", (
+            f"cleanup.config mismatch: {cleanup.get('config')!r}"
+        )
+        assert "contexts" in cleanup, f"Missing 'contexts' in cleanup: {cleanup}"
+
+        messages = payload.get("messages", [])
+        assert messages, f"Expected non-empty messages list, got: {messages}"
+        assert messages[0].get("level") == "ok", (
+            f"Unexpected message level: {messages[0].get('level')!r}"
+        )
+        assert "Actor removed" in messages[0].get("text", ""), (
+            f"Expected 'Actor removed' in message text: {messages[0].get('text')!r}"
+        )
+
+        print("actor-remove-json-format-ok")
+
+    finally:
+        cleanup_workspace(workspace)
+
+
+def main() -> None:
+    command = sys.argv[1] if len(sys.argv) > 1 else "remove-json"
+    dispatch: dict[str, object] = {
+        "remove-json": test_remove_format_json,
+    }
+    handler = dispatch.get(command)
+    if handler is None:
+        print(f"Unknown command: {command}", file=sys.stderr)
+        sys.exit(1)
+    assert callable(handler)
+    handler()
+
+
+if __name__ == "__main__":
+    main()

src/cleveragents/acms/__init__.py

@@ -5,12 +5,22 @@ technology-specific vocabulary extensions, and the DetailLevelMap
 inheritance mechanism for resolving named detail levels across the
 ontology hierarchy (Layer 3 -> Layer 2 -> Layer 1 -> Layer 0).
 
+Also provides the ACMS index data model and file traversal engine for
+indexing large projects.
+
 Based on ``docs/specification.md`` ~lines 42333-42422, 44405-44420.
 """
 
 from __future__ import annotations
 
 from cleveragents.acms import uko as _uko
+from cleveragents.acms.index import (
+    ACMSIndex,
+    FileTraversalEngine,
+    FileType,
+    IndexEntry,
+    TierLevel,
+)
 from cleveragents.acms.uko import (
     CODE_DETAIL_LEVEL_MAP,
     FUNC_DETAIL_LEVEL_MAP,
@@ -62,6 +72,14 @@ from cleveragents.acms.uko import (
     resolve_detail_level,
 )
 
-# Re-export everything published by the ``uko`` sub-package so the two
-# ``__all__`` lists stay in sync automatically.
-__all__: list[str] = list(_uko.__all__)
+# Combine exports from both uko and index modules
+_uko_exports = list(_uko.__all__)
+_index_exports = [
+    "ACMSIndex",
+    "FileTraversalEngine",
+    "FileType",
+    "IndexEntry",
+    "TierLevel",
+]
+
+__all__: list[str] = _uko_exports + _index_exports

src/cleveragents/acms/index.py

@@ -0,0 +1,412 @@
+"""ACMS Index Data Model and File Traversal Engine.
+
+Provides the foundational data model for indexed context entries and a
+file traversal engine that can handle 10,000+ files without timeout using
+chunked processing.
+
+Based on issue #9579 and ``docs/specification.md`` ~lines 44405-44420.
+"""
+
+from __future__ import annotations
+
+from collections.abc import Iterator
+from dataclasses import dataclass, field
+from datetime import datetime
+from enum import StrEnum
+from pathlib import Path
+
+
+class FileType(StrEnum):
+    """File type enumeration for index entries."""
+
+    PYTHON = "python"
+    JAVASCRIPT = "javascript"
+    TYPESCRIPT = "typescript"
+    JAVA = "java"
+    RUST = "rust"
+    MARKDOWN = "markdown"
+    TEXT = "text"
+    JSON = "json"
+    YAML = "yaml"
+    XML = "xml"
+    OTHER = "other"
+
+
+class TierLevel(StrEnum):
+    """Tier assignment levels for context entries.
+
+    Aligns with the hot/warm/cold storage tier vocabulary used throughout
+    the ACMS specification and milestone v3.4.0 acceptance criteria.
+    """
+
+    HOT = "hot"  # Core/essential
+    WARM = "warm"  # Important
+    COLD = "cold"  # Supporting
+    ARCHIVE = "archive"  # Reference
+
+
+@dataclass
+class IndexEntry:
+    """Represents a single indexed context entry.
+
+    Attributes:
+        path: Absolute or relative file path
+        file_type: Type of file (from FileType enum)
+        size_bytes: File size in bytes
+        created_at: File creation timestamp
+        modified_at: File modification timestamp
+        tags: Set of tags for categorization
+        tier: Tier assignment level
+        metadata: Additional metadata dictionary
+    """
+
+    path: str
+    file_type: FileType
+    size_bytes: int
+    created_at: datetime
+    modified_at: datetime
+    tags: set[str] = field(default_factory=set)
+    tier: TierLevel = TierLevel.COLD
+    metadata: dict[str, str] = field(default_factory=dict)
+
+    def add_tag(self, tag: str) -> None:
+        """Add a tag to this entry.
+
+        Args:
+            tag: Tag string to add. Must be non-empty.
+
+        Raises:
+            ValueError: If tag is empty or whitespace-only.
+        """
+        if not tag or not tag.strip():
+            raise ValueError("tag must be a non-empty string")
+        self.tags.add(tag)
+
+    def remove_tag(self, tag: str) -> None:
+        """Remove a tag from this entry.
+
+        Args:
+            tag: Tag string to remove. Must be non-empty.
+
+        Raises:
+            ValueError: If tag is empty or whitespace-only.
+        """
+        if not tag or not tag.strip():
+            raise ValueError("tag must be a non-empty string")
+        self.tags.discard(tag)
+
+    def has_tag(self, tag: str) -> bool:
+        """Check if entry has a specific tag."""
+        return tag in self.tags
+
+    def set_tier(self, tier: TierLevel) -> None:
+        """Set the tier level for this entry."""
+        self.tier = tier
+
+
+@dataclass
+class ACMSIndex:
+    """ACMS Index for storing and querying indexed context entries.
+
+    Provides storage and query capabilities for indexed files with support
+    for filtering by path, tags, type, and recency.
+
+    Attributes:
+        entries: Dictionary mapping file paths to IndexEntry objects
+    """
+
+    entries: dict[str, IndexEntry] = field(default_factory=dict)
+
+    def add_entry(self, entry: IndexEntry) -> None:
+        """Add an index entry to the index.
+
+        Args:
+            entry: The IndexEntry to add. Must not be None.
+
+        Raises:
+            TypeError: If entry is not an IndexEntry instance.
+        """
+        if not isinstance(entry, IndexEntry):
+            raise TypeError(
+                f"entry must be an IndexEntry instance, got {type(entry).__name__}"
+            )
+        self.entries[entry.path] = entry
+
+    def remove_entry(self, path: str) -> bool:
+        """Remove an entry by path. Returns True if removed, False if not found."""
+        if path in self.entries:
+            del self.entries[path]
+            return True
+        return False
+
+    def get_entry(self, path: str) -> IndexEntry | None:
+        """Get an entry by path."""
+        return self.entries.get(path)
+
+    def query_by_path(self, path_pattern: str) -> list[IndexEntry]:
+        """Query entries by path pattern (substring match).
+
+        Args:
+            path_pattern: Substring pattern to match against entry paths.
+                Must be non-empty.
+
+        Raises:
+            ValueError: If path_pattern is empty or whitespace-only.
+        """
+        if not path_pattern or not path_pattern.strip():
+            raise ValueError("path_pattern must be a non-empty string")
+        return [entry for entry in self.entries.values() if path_pattern in entry.path]
+
+    def query_by_tag(self, tag: str) -> list[IndexEntry]:
+        """Query entries by tag."""
+        return [entry for entry in self.entries.values() if entry.has_tag(tag)]
+
+    def query_by_type(self, file_type: FileType) -> list[IndexEntry]:
+        """Query entries by file type."""
+        return [
+            entry for entry in self.entries.values() if entry.file_type == file_type
+        ]
+
+    def query_by_tier(self, tier: TierLevel) -> list[IndexEntry]:
+        """Query entries by tier level."""
+        return [entry for entry in self.entries.values() if entry.tier == tier]
+
+    def query_by_recency(
+        self, after: datetime, before: datetime | None = None
+    ) -> list[IndexEntry]:
+        """Query entries by modification recency.
+
+        Args:
+            after: Return entries modified after this datetime.
+            before: Return entries modified before this datetime (optional).
+                When both are provided, before must be >= after.
+
+        Returns:
+            List of entries matching the recency criteria.
+
+        Raises:
+            ValueError: If before is earlier than after when both are provided.
+        """
+        if before is not None and before < after:
+            raise ValueError(
+                f"before ({before}) must be >= after ({after}) when both are provided"
+            )
+        results = [
+            entry for entry in self.entries.values() if entry.modified_at >= after
+        ]
+        if before:
+            results = [entry for entry in results if entry.modified_at <= before]
+        return results
+
+    def query_combined(
+        self,
+        path_pattern: str | None = None,
+        tags: set[str] | None = None,
+        file_type: FileType | None = None,
+        tier: TierLevel | None = None,
+        after: datetime | None = None,
+    ) -> list[IndexEntry]:
+        """Query entries with multiple filters (AND logic).
+
+        Args:
+            path_pattern: Filter by path pattern
+            tags: Filter by any of these tags
+            file_type: Filter by file type
+            tier: Filter by tier level
+            after: Filter by modification date
+
+        Returns:
+            List of entries matching all specified criteria
+        """
+        results = list(self.entries.values())
+
+        if path_pattern:
+            results = [entry for entry in results if path_pattern in entry.path]
+
+        if tags:
+            results = [
+                entry for entry in results if any(tag in entry.tags for tag in tags)
+            ]
+
+        if file_type:
+            results = [entry for entry in results if entry.file_type == file_type]
+
+        if tier:
+            results = [entry for entry in results if entry.tier == tier]
+
+        if after:
+            results = [entry for entry in results if entry.modified_at >= after]
+
+        return results
+
+    def get_all_entries(self) -> list[IndexEntry]:
+        """Get all entries in the index."""
+        return list(self.entries.values())
+
+    def get_entry_count(self) -> int:
+        """Get the total number of entries in the index."""
+        return len(self.entries)
+
+
+class FileTraversalEngine:
+    """Engine for traversing and indexing files in large projects.
+
+    Handles 10,000+ files without timeout using chunked processing to
+    prevent memory exhaustion.
+
+    Attributes:
+        chunk_size: Number of files to process in each chunk
+        index: The ACMS index to populate
+    """
+
+    def __init__(self, chunk_size: int = 100, index: ACMSIndex | None = None) -> None:
+        """Initialize the traversal engine.
+
+        Args:
+            chunk_size: Number of files to process per chunk (default: 100).
+                Must be a positive integer.
+            index: Optional pre-populated ACMSIndex to use. If None, a new
+                empty index is created (supports Dependency Inversion).
+
+        Raises:
+            ValueError: If chunk_size is not a positive integer.
+        """
+        if chunk_size <= 0:
+            raise ValueError(f"chunk_size must be a positive integer, got {chunk_size}")
+        self.chunk_size = chunk_size
+        self.index = index if index is not None else ACMSIndex()
+
+    def _get_file_type(self, file_path: Path) -> FileType:
+        """Determine file type from extension."""
+        suffix = file_path.suffix.lower()
+        type_map = {
+            ".py": FileType.PYTHON,
+            ".js": FileType.JAVASCRIPT,
+            ".ts": FileType.TYPESCRIPT,
+            ".tsx": FileType.TYPESCRIPT,
+            ".java": FileType.JAVA,
+            ".rs": FileType.RUST,
+            ".md": FileType.MARKDOWN,
+            ".txt": FileType.TEXT,
+            ".json": FileType.JSON,
+            ".yaml": FileType.YAML,
+            ".yml": FileType.YAML,
+            ".xml": FileType.XML,
+        }
+        return type_map.get(suffix, FileType.OTHER)
+
+    def _create_index_entry(self, file_path: Path) -> IndexEntry | None:
+        """Create an index entry from a file path.
+
+        Args:
+            file_path: Path to the file
+
+        Returns:
+            IndexEntry if successful, None if file cannot be read
+        """
+        try:
+            stat = file_path.stat()
+            return IndexEntry(
+                path=str(file_path),
+                file_type=self._get_file_type(file_path),
+                size_bytes=stat.st_size,
+                created_at=datetime.fromtimestamp(stat.st_ctime),
+                modified_at=datetime.fromtimestamp(stat.st_mtime),
+            )
+        except (OSError, ValueError):
+            return None
+
+    def _traverse_directory(
+        self,
+        root_path: Path,
+    ) -> Iterator[Path]:
+        """Recursively traverse directory and yield file paths.
+
+        Args:
+            root_path: Root directory to traverse
+
+        Yields:
+            Path objects for each file found
+        """
+        try:
+            for item in root_path.iterdir():
+                if item.is_file():
+                    yield item
+                elif item.is_dir():
+                    # Recursively traverse subdirectories
+                    yield from self._traverse_directory(item)
+        except (OSError, PermissionError):
+            # Skip directories we can't read
+            pass
+
+    def traverse_and_index(
+        self,
+        root_path: str | Path,
+        exclude_patterns: list[str] | None = None,
+    ) -> ACMSIndex:
+        """Traverse a directory and index all files.
+
+        Uses chunked processing to handle large projects without timeout
+        or memory exhaustion.
+
+        Args:
+            root_path: Root directory to traverse
+            exclude_patterns: List of patterns to exclude
+                (e.g., ['.git', '__pycache__'])
+
+        Returns:
+            Populated ACMSIndex
+        """
+        root = Path(root_path)
+        if not root.exists():
+            raise ValueError(f"Path does not exist: {root_path}")
+
+        exclude_patterns = exclude_patterns or []
+        chunk: list[IndexEntry] = []
+
+        for file_path in self._traverse_directory(root):
+            # Check if file matches any exclude pattern
+            if any(pattern in str(file_path) for pattern in exclude_patterns):
+                continue
+
+            # Create index entry
+            entry = self._create_index_entry(file_path)
+            if entry:
+                chunk.append(entry)
+
+            # Process chunk when it reaches the size limit
+            if len(chunk) >= self.chunk_size:
+                self._process_chunk(chunk)
+                chunk = []
+
+        # Process remaining entries
+        if chunk:
+            self._process_chunk(chunk)
+
+        return self.index
+
+    def _process_chunk(self, chunk: list[IndexEntry]) -> None:
+        """Process a chunk of index entries.
+
+        Args:
+            chunk: List of IndexEntry objects to add to the index
+        """
+        for entry in chunk:
+            self.index.add_entry(entry)
+
+    def get_index(self) -> ACMSIndex:
+        """Get the current index."""
+        return self.index
+
+    def reset_index(self) -> None:
+        """Reset the index to empty state."""
+        self.index = ACMSIndex()
+
+
+__all__ = [
+    "ACMSIndex",
+    "FileTraversalEngine",
+    "FileType",
+    "IndexEntry",
+    "TierLevel",
+]

src/cleveragents/actor/compiler.py

@@ -137,7 +137,7 @@ def _map_node(node: NodeDefinition) -> lg_nodes.NodeConfig:
             config.get("function") if node.type == NodeType.CONDITIONAL else None
         ),
         tools=config.get("tools", []) if node.type == NodeType.TOOL else [],
-        subgraph=(config.get("actor_ref") if node.type == NodeType.SUBGRAPH else None),
+        subgraph=(node.actor_ref if node.type == NodeType.SUBGRAPH else None),
         metadata=dict(config),
     )
 
@@ -187,6 +187,11 @@ def _detect_subgraph_cycles(
     """Detect cycles in subgraph references across actors.
 
     Returns list of actor names forming a cycle, or empty list.
+
+    The actor reference is read from the top-level actor_ref field on
+    NodeDefinition, not from node.config.  Reading from config
+    would always return an empty string because actor_ref is stored as a
+    first-class field on the schema model.
     """
     if resolver is None:
         return []
@@ -194,7 +199,7 @@ def _detect_subgraph_cycles(
     for node in route_nodes:
         if node.type != NodeType.SUBGRAPH:
             continue
-        ref_name = node.config.get("actor_ref", "")
+        ref_name = node.actor_ref or ""
         if not ref_name:
             continue
         if ref_name in visited:
@@ -290,7 +295,7 @@ def compile_actor(
         all_lsp_bindings.extend(bindings)
 
         if node_def.type == NodeType.SUBGRAPH:
-            ref = node_def.config.get("actor_ref", "")
+            ref = node_def.actor_ref or ""
             if ref:
                 subgraph_refs[node_def.id] = ref
 

src/cleveragents/cli/commands/actor.py

@@ -816,12 +816,28 @@ def update(
 
 
 @app.command()
-def remove(name: Annotated[str, typer.Argument(help="Actor name to remove")]) -> None:
+def remove(
+    name: Annotated[str, typer.Argument(help="Actor name to remove")],
+    fmt: Annotated[
+        str,
+        typer.Option("--format", "-f", help=_FORMAT_HELP),
+    ] = OutputFormat.RICH.value,
+) -> None:
     """Remove a custom actor.
 
     Specify the namespaced name (e.g. ``local/my-actor``).
     """
 
+    # Validate --format argument first; fail fast before any side effects.
+    fmt_value = fmt.lower()
+    _valid_formats = {f.value for f in OutputFormat}
+    if fmt_value not in _valid_formats:
+        raise typer.BadParameter(
+            f"Invalid format {fmt!r}. "
+            f"Supported values: {', '.join(sorted(_valid_formats))}",
+            param_hint="'--format'",
+        )
+
     service, registry = _get_services()
     try:
         # Get actor details before removal for display
@@ -844,6 +860,40 @@ def remove(name: Annotated[str, typer.Argument(help="Actor name to remove")]) ->
         else:
             service.remove_actor(name)
 
+        command_name = f"agents actor remove {name}"
+        payload = {
+            "actor_removed": {
+                "name": name,
+                "provider": actor_provider,
+                "model": actor_model,
+            },
+            "impact": {
+                "sessions": session_count,
+                "active_plans": active_plan_count,
+                "actions_referencing": action_count,
+            },
+            "cleanup": {
+                "config": "kept on disk",
+                # NOTE: context-cleanup count is deferred; always 0 for now.
+                # Follow-up: implement dynamic orphaned-context detection.
+                "contexts": "0 orphaned",
+            },
+        }
+        messages = [{"level": "ok", "text": "Actor removed"}]
+
+        if fmt_value != OutputFormat.RICH.value:
+            rendered = format_output(
+                payload,
+                fmt_value,
+                command=command_name,
+                status="ok",
+                exit_code=0,
+                messages=messages,
+            )
+            if rendered:
+                console.print(rendered)
+            return
+
         # Display Actor Removed panel
         actor_info = (
             f"[cyan bold]Name:[/cyan bold] {name}\n"

src/cleveragents/cli/commands/session.py

@@ -151,8 +151,8 @@ def _session_list_dict(sessions: list[Session]) -> dict[str, Any]:
     # Find most recent and oldest sessions
     if sessions:
         sorted_sessions = sorted(sessions, key=lambda x: x.updated_at, reverse=True)
-        most_recent = sorted_sessions[0].name or sorted_sessions[0].session_id[:8]
-        oldest = sorted_sessions[-1].name or sorted_sessions[-1].session_id[:8]
+        most_recent = sorted_sessions[0].name or sorted_sessions[0].session_id
+        oldest = sorted_sessions[-1].name or sorted_sessions[-1].session_id
     else:
         most_recent = None
         oldest = None
@@ -347,7 +347,7 @@ def list_sessions(
 
     for s in sessions:
         table.add_row(
-            s.session_id[:8],  # Truncate ID for readability
+            s.session_id,  # Full ULID for copy-paste compatibility with session tell
             s.name or "(unnamed)",
             s.actor_name or "(none)",
             str(s.message_count),

src/cleveragents/infrastructure/database/llm_trace_repository.py

@@ -31,6 +31,17 @@ class LLMTraceRepository:
 
     Uses the session-factory pattern: each public method obtains a
     session from the factory.  Callers are responsible for commit.
+
+    When ``save()`` is called with an explicit ``session`` argument the
+    repository operates in *UnitOfWork mode*: it flushes the change into
+    the caller's transaction but does **not** commit or close the session.
+    The caller (or the enclosing ``UnitOfWork``) is responsible for the
+    final commit.
+
+    When ``save()`` is called without an explicit ``session`` argument the
+    repository operates in *standalone mode*: it creates its own session
+    from the factory, flushes, commits, and closes the session so that the
+    trace is durably persisted even outside a ``UnitOfWork``.
     """
 
     def __init__(
@@ -46,16 +57,26 @@ class LLMTraceRepository:
         return self._sf()
 
     @database_retry
-    def save(self, trace: LLMTrace) -> None:
+    def save(self, trace: LLMTrace, session: Session | None = None) -> None:
         """Persist a single ``LLMTrace`` row.
 
         Args:
-            trace: The trace to persist.
+            trace: The trace to persist.  Must not be ``None``.
+            session: Optional external SQLAlchemy session.  When provided
+                the repository flushes into the caller's transaction and
+                does **not** commit or close the session (UnitOfWork mode).
+                When omitted the repository creates its own session, commits,
+                and closes it (standalone mode).
 
         Raises:
+            ValueError: If ``trace`` is ``None``.
             DatabaseError: On unrecoverable persistence failure.
         """
-        session = self._session()
+        if trace is None:
+            raise ValueError("trace must not be None")
+
+        own_session = session is None
+        s: Session = self._session() if own_session else session
         try:
             model = LLMTraceModel(
                 trace_id=trace.trace_id,
@@ -77,11 +98,16 @@ class LLMTraceRepository:
                 error=trace.error,
                 timestamp=trace.timestamp.isoformat(),
             )
-            session.add(model)
-            session.flush()
+            s.add(model)
+            s.flush()
+            if own_session:
+                s.commit()
         except (SQLAlchemyDatabaseError, OperationalError) as exc:
-            session.rollback()
+            s.rollback()
             raise DatabaseError(f"Failed to save LLM trace: {exc}") from exc
+        finally:
+            if own_session:
+                s.close()
 
     @database_retry
     def get(self, trace_id: str) -> LLMTrace | None:

src/cleveragents/infrastructure/database/migration_runner.py

@@ -151,10 +151,22 @@ class MigrationRunner:
     def get_current_revision(self) -> str | None:
         """Get the current migration revision of the database.
 
+        For SQLite databases, the engine is created with
+        ``connect_args={"check_same_thread": False}`` so that this method
+        can be safely called from any thread — including background threads
+        used in async startup flows.  This is consistent with the pattern
+        used in :meth:`init_or_upgrade`.
+
         Returns:
             Current revision ID or None if no migrations have been applied
         """
-        engine = create_engine(self.database_url)
+        if self.database_url.startswith("sqlite"):
+            engine = create_engine(
+                self.database_url,
+                connect_args={"check_same_thread": False},
+            )
+        else:
+            engine = create_engine(self.database_url)
         with engine.connect() as connection:
             context = MigrationContext.configure(connection)
             return context.get_current_revision()

src/cleveragents/tui/session_store.py

@@ -0,0 +1,269 @@
+"""SQLite-based session persistence for TUI."""
+
+from __future__ import annotations
+
+import sqlite3
+import threading
+from dataclasses import dataclass, field
+from datetime import UTC, datetime
+from pathlib import Path
+
+ALLOWED_STATES = ("idle", "working", "awaiting_input")
+
+
+@dataclass(slots=True)
+class SessionRecord:
+    """A persisted TUI session record."""
+
+    session_id: str
+    name: str
+    state: str  # "idle", "working", "awaiting_input"
+    created_at: datetime
+    updated_at: datetime
+    transcript: str = field(default="[]")
+
+
+def _utcnow() -> datetime:
+    """Return the current UTC timestamp (timezone-aware)."""
+    return datetime.now(UTC)
+
+
+class SessionStore:
+    """SQLite-based session persistence for TUI sessions.
+
+    Provides thread-safe CRUD operations for persisting TUI session state,
+    names, and timestamps to a local SQLite database file.
+    """
+
+    def __init__(self, db_path: Path | None = None) -> None:
+        """Initialize the session store.
+
+        Args:
+            db_path: Path to SQLite database. Defaults to
+                ``~/.local/state/cleveragents/tui.db``
+        """
+        if db_path is None:
+            state_dir = Path.home() / ".local" / "state" / "cleveragents"
+            db_path = state_dir / "tui.db"
+
+        self.db_path: Path = Path(db_path)
+        self._lock = threading.Lock()
+        self._init_db()
+
+    def _init_db(self) -> None:
+        """Initialize database schema."""
+        self.db_path.parent.mkdir(parents=True, exist_ok=True)
+        conn = sqlite3.connect(
+            str(self.db_path),
+            check_same_thread=False,
+        )
+        try:
+            conn.execute("""
+                CREATE TABLE IF NOT EXISTS sessions (
+                    session_id TEXT PRIMARY KEY,
+                    name TEXT NOT NULL,
+                    state TEXT NOT NULL DEFAULT 'idle',
+                    created_at TEXT NOT NULL,
+                    updated_at TEXT NOT NULL,
+                    transcript TEXT NOT NULL DEFAULT '[]'
+                )
+            """)
+            conn.commit()
+        finally:
+            conn.close()
+
+    def create_session(
+        self, session_id: str, name: str, state: str = "idle"
+    ) -> SessionRecord:
+        """Create a new session.
+
+        Args:
+            session_id: Unique session identifier.
+            name: Human-readable session name.
+            state: Initial state (one of *idle*, *working*,
+                *awaiting_input*). Defaults to ``"idle"``.
+
+        Returns:
+            The created :class:`SessionRecord`.
+
+        Raises:
+            ValueError: If *state* is not one of the allowed values.
+        """
+        if state not in ALLOWED_STATES:
+            raise ValueError(
+                f"Invalid state '{state}'. Must be one of {ALLOWED_STATES}."
+            )
+
+        now = _utcnow()
+
+        with self._lock:
+            conn = sqlite3.connect(
+                str(self.db_path),
+                check_same_thread=False,
+            )
+            try:
+                conn.execute(
+                    """
+                    INSERT INTO sessions
+                    (session_id, name, state, created_at, updated_at, transcript)
+                    VALUES (?, ?, ?, ?, ?, ?)
+                    """,
+                    (
+                        session_id,
+                        name,
+                        state,
+                        now.isoformat(),
+                        now.isoformat(),
+                        "[]",
+                    ),
+                )
+                conn.commit()
+            finally:
+                conn.close()
+
+        return SessionRecord(
+            session_id=session_id,
+            name=name,
+            state=state,
+            created_at=now,
+            updated_at=now,
+            transcript="[]",
+        )
+
+    def get_session(self, session_id: str) -> SessionRecord | None:
+        """Retrieve a session by ID.
+
+        Args:
+            session_id: The session identifier.
+
+        Returns:
+            The :class:`SessionRecord` if found, ``None`` otherwise.
+        """
+        with self._lock:
+            conn = sqlite3.connect(
+                str(self.db_path),
+                check_same_thread=False,
+            )
+            try:
+                cursor = conn.execute(
+                    """
+                    SELECT session_id, name, state, created_at, updated_at,
+                           transcript
+                    FROM sessions WHERE session_id = ?
+                    """,
+                    (session_id,),
+                )
+                row = cursor.fetchone()
+            finally:
+                conn.close()
+
+        if row is None:
+            return None
+
+        return SessionRecord(
+            session_id=row[0],
+            name=row[1],
+            state=row[2],
+            created_at=datetime.fromisoformat(row[3]),
+            updated_at=datetime.fromisoformat(row[4]),
+            transcript=row[5],
+        )
+
+    def list_sessions(self) -> list[SessionRecord]:
+        """List all stored sessions.
+
+        Returns:
+            List of :class:`SessionRecord` objects ordered by creation time
+            (oldest first).
+        """
+        with self._lock:
+            conn = sqlite3.connect(
+                str(self.db_path),
+                check_same_thread=False,
+            )
+            try:
+                cursor = conn.execute(
+                    """
+                    SELECT session_id, name, state, created_at, updated_at,
+                           transcript
+                    FROM sessions ORDER BY created_at ASC
+                    """
+                )
+                rows = cursor.fetchall()
+            finally:
+                conn.close()
+
+        return [
+            SessionRecord(
+                session_id=row[0],
+                name=row[1],
+                state=row[2],
+                created_at=datetime.fromisoformat(row[3]),
+                updated_at=datetime.fromisoformat(row[4]),
+                transcript=row[5],
+            )
+            for row in rows
+        ]
+
+    def update_session_state(self, session_id: str, state: str) -> None:
+        """Update a session's state.
+
+        Args:
+            session_id: The session identifier.
+            state: New state value (one of *idle*, *working*,
+                *awaiting_input*).
+
+        Raises:
+            ValueError: If *state* is not one of the allowed values.
+        """
+        if state not in ALLOWED_STATES:
+            raise ValueError(
+                f"Invalid state '{state}'. Must be one of {ALLOWED_STATES}."
+            )
+
+        now = _utcnow()
+
+        with self._lock:
+            conn = sqlite3.connect(
+                str(self.db_path),
+                check_same_thread=False,
+            )
+            try:
+                conn.execute(
+                    "UPDATE sessions SET "
+                    "state = ?, updated_at = ? "
+                    "WHERE session_id = ?",
+                    (state, now.isoformat(), session_id),
+                )
+                conn.commit()
+            finally:
+                conn.close()
+
+    def delete_session(self, session_id: str) -> None:
+        """Delete a session.
+
+        Args:
+            session_id: The session identifier.
+        """
+        with self._lock:
+            conn = sqlite3.connect(
+                str(self.db_path),
+                check_same_thread=False,
+            )
+            try:
+                conn.execute(
+                    "DELETE FROM sessions WHERE session_id = ?",
+                    (session_id,),
+                )
+                conn.commit()
+            finally:
+                conn.close()
+
+    def close(self) -> None:
+        """Close the database connection.
+
+        SQLite connections are managed per-operation with
+        ``check_same_thread=False``; no long-lived handle exists.
+        This method is retained for API completeness and is a no-op.
+        """
+        pass

src/cleveragents/tui/widgets/__init__.py

@@ -10,6 +10,7 @@ from cleveragents.tui.widgets.permission_question import (
 from cleveragents.tui.widgets.persona_bar import PersonaBar
 from cleveragents.tui.widgets.prompt import PromptInput, PromptSubmitted
 from cleveragents.tui.widgets.reference_picker import ReferencePickerOverlay
+from cleveragents.tui.widgets.session_tab_bar import SessionTabBar
 from cleveragents.tui.widgets.slash_command_overlay import SlashCommandOverlay
 from cleveragents.tui.widgets.thought_block import ThoughtBlockWidget
 
@@ -22,6 +23,7 @@ __all__ = [
     "PromptInput",
     "PromptSubmitted",
     "ReferencePickerOverlay",
+    "SessionTabBar",
     "SlashCommandOverlay",
     "ThoughtBlockWidget",
     "render_permission_question",

src/cleveragents/tui/widgets/session_tab_bar.py

@@ -0,0 +1,220 @@
+"""Session tab bar widget for multi-session TUI."""
+
+from __future__ import annotations
+
+import contextlib
+import importlib
+from collections.abc import Callable
+from typing import Any, ClassVar
+
+
+def _load_static_base() -> type[Any]:
+    """Load the ``Static`` base class from *Textual* or return a fallback.
+
+    Returns:
+        A class that supports ``update(text)`` and has a ``display`` bool.
+    """
+    try:
+        tv = importlib.import_module("textual.widgets")
+        static_cls = getattr(tv, "Static", None)
+        if static_cls is not None:
+            return static_cls
+    except Exception:  # pragma: no cover - optional dependency
+        pass
+
+    class _FallbackStatic:
+        """Fallback *Static* widget when Textual is not available."""
+
+        def __init__(self, *args: Any, **kwargs: Any) -> None:
+            """Initialize the fallback widget."""
+            object.__setattr__(self, "_text", "")
+            object.__setattr__(self, "display", True)
+
+        def update(self, text: str) -> None:
+            """Update the widget text content."""
+            object.__setattr__(self, "_text", text)
+
+    return _FallbackStatic
+
+
+_StaticBase = _load_static_base()
+
+
+class SessionTabBar(_StaticBase):
+    """Widget displaying session tabs with state indicators.
+
+    Shows tabs for each session with state indicators:
+
+    * ``⌛`` for *working* state (hourglass)
+    * ``>`` for *awaiting_input* state (prompt arrow, U+276F)
+    * empty string for *idle* state (no indicator)
+
+    Automatically hidden when there is only one (or zero) sessions in the
+    store.
+
+    When a session is marked as active it is wrapped in square brackets,
+    e.g. ``[> Main Session]``.
+    """
+
+    DEFAULT_CSS: ClassVar[str] = """
+    SessionTabBar {
+        height: 1;
+        background: $panel;
+        border: solid $primary;
+        border-top: none;
+        border-left: none;
+        border-right: none;
+    }
+    """
+
+    _CallbackDict = dict[str, Callable[..., Any]]
+
+    def __init__(self, *, id: str | None = None, classes: str | None = None) -> None:
+        """Initialize the session tab bar.
+
+        Args:
+            id: Widget ID for *Textual* targeting.
+            classes: CSS classes applied to this widget.
+        """
+        super().__init__(id=id, classes=classes)
+        self._sessions: list[dict[str, str]] = []
+        self._active_session_id: str | None = None
+        self._current_index: int = 0
+        self._callbacks: SessionTabBar._CallbackDict = {}
+        self._current_text: str = ""
+
+    def set_on_navigation(self, callbacks: _CallbackDict) -> None:
+        """Register navigation action callbacks.
+
+        Args:
+            callbacks: Dict of 'prev', 'next', 'new', 'close',
+                'session_selected', 'to_sessions_screen'.
+        """
+        self._callbacks.update(callbacks)
+
+    def navigate_prev(self) -> None:
+        """Go to the previous session tab."""
+        num = len(self._sessions)
+        if num == 0:
+            return
+        self._current_index = (self._current_index - 1) % num
+        cb = self._callbacks.get("session_selected")
+        if cb is not None:
+            cb(self._sessions[self._current_index])
+
+    def navigate_next(self) -> None:
+        """Go to the next session tab."""
+        num = len(self._sessions)
+        if num == 0:
+            return
+        self._current_index = (self._current_index + 1) % num
+        cb = self._callbacks.get("session_selected")
+        if cb is not None:
+            cb(self._sessions[self._current_index])
+
+    def create_new_session(self) -> None:
+        """Request creation of a new session."""
+        cb = self._callbacks.get("new")
+        if cb is not None:
+            cb()
+
+    def close_current_session(self) -> None:
+        """Close the currently selected session tab."""
+        num = len(self._sessions)
+        if num == 0:
+            return
+        removed = self._sessions.pop(self._current_index)
+        if self._current_index >= len(self._sessions):
+            self._current_index = max(0, len(self._sessions) - 1)
+        cb = self._callbacks.get("close")
+        if cb is not None:
+            cb(removed)
+
+    def jump_to_session(self, index: int) -> None:
+        """Jump to a session by 1-based index key (1-9).
+
+        Args:
+            index: The tab number (1-based). Clamped to valid range.
+        """
+        if index < 1 or index > len(self._sessions):
+            return
+        self._current_index = index - 1
+        cb = self._callbacks.get("session_selected")
+        if cb is not None:
+            cb(self._sessions[self._current_index])
+
+    def switch_to_sessions_screen(self) -> None:
+        """Request navigation to the Sessions screen."""
+        cb = self._callbacks.get("to_sessions_screen")
+        if cb is not None:
+            cb()
+
+    def set_sessions(
+        self,
+        sessions: list[dict[str, str]],
+        active_session_id: str | None = None,
+    ) -> None:
+        """Update the displayed sessions.
+
+        Args:
+            sessions: List of session dicts containing keys
+                ``session_id``, ``name``, and ``state``.
+            active_session_id: ID of the currently active session (None
+                means no session is highlighted).
+        """
+        self._sessions = sessions
+        self._active_session_id = active_session_id
+        num = len(sessions)
+        if self._current_index >= num:
+            self._current_index = max(0, num - 1)
+        self._render()
+
+    def _render(self) -> None:
+        """Render the tab bar content based on current session list."""
+        if len(self._sessions) <= 1:
+            object.__setattr__(self, "display", False)
+            with contextlib.suppress(Exception):
+                self.visible = False
+            self._current_text = ""
+            object.__setattr__(self, "_text", "")
+            return
+
+        object.__setattr__(self, "display", True)
+        with contextlib.suppress(Exception):
+            self.visible = True
+        tabs: list[str] = []
+
+        for session in self._sessions:
+            session_id = session.get("session_id", "")
+            name = session.get("name", "")
+            state = session.get("state", "idle")
+
+            # State indicator prefix
+            indicator = self._get_state_indicator(state)
+            tab_text = f"{indicator} {name}".strip()
+
+            # Mark active session with brackets
+            if session_id == self._active_session_id:
+                tab_text = f"[{tab_text}]"
+
+            tabs.append(tab_text)
+
+        content = " | ".join(tabs)
+        self._current_text = content
+        getattr(self, "update", lambda t: None)(content)
+
+    @staticmethod
+    def _get_state_indicator(state: str) -> str:
+        """Return the Unicode state indicator for a given session state.
+
+        Args:
+            state: One of ``idle``, ``working``, or ``awaiting_input``.
+
+        Returns:
+            The indicator character (or empty string for idle).
+        """
+        if state == "working":
+            return "\u231b"  # ⌛ hourglass
+        elif state == "awaiting_input":
+            return "\u276f"  # > prompt arrow (per issue #5330 spec)
+        return ""