test(kanban): cover review-claimed rejection tool path

[sahilm-ai]

Summary

• Cover the public kanban_reject tool path after claim_review_task transitions a review card to running.
• Assert the reviewer run closes as rejected, the task returns to ready, and the rejection event remains attached to that run.
• The DB implementation is already present on fork main; this closes the sdlc-review integration-regression gap.

Verification

• scripts/run_tests.sh tests/tools/test_kanban_tools.py -q — 111 passed.
• ruff check tests/tools/test_kanban_tools.py — passed.
• ty check tests/tools/test_kanban_tools.py — reports 68 pre-existing diagnostics elsewhere in the file; none in this test.
• Full scripts/run_tests.sh is running separately. Its .venv environment has existing ACP import failures (ModuleNotFoundError: acp) before this change's suite.

Summary by CodeRabbit

• Tests
  • Added coverage for rejecting tasks that were claimed while in review.
  • Verifies rejected tasks return to ready status, release active claims, and record the expected run and event details.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 7b5e8605-9d77-4e55-9d47-72b573765a69

📥 Commits

Reviewing files that changed from the base of the PR and between 1c526d6 and 8205d17.

📒 Files selected for processing (1)

• tests/tools/test_kanban_tools.py

---

📝 Walkthrough

Walkthrough

Adds coverage for rejecting a task moved from review to running through review claiming, verifying reset state and recorded rejection audit data.

Changes

Rejection handling coverage

Layer / File(s)	Summary
Review-claimed rejection validation tests/tools/test_kanban_tools.py	Adds a test covering _handle_reject, task rollback to ready, cleared claim fields, and expected task_runs and task_events records.

Estimated code review effort: 2 (Simple) | ~10 minutes

Possibly related PRs

• sahilm-ti/hermes-agent#10: Implements the related rejection path for running tasks originating from review.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately summarizes the main change: a kanban test covering the review-claimed rejection path.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/t-361a29a5-reject-review-claimed

⚔️ Resolve merge conflicts

• Resolve merge conflict in branch fix/t-361a29a5-reject-review-claimed

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[sahilm-ai]

auto-review: approved.

Matrix checks (U1–U5, C1–C5)

• U1 in-scope files: PASS — one focused regression test for the review-claimed rejection path.
• U2 out-of-scope deletions: PASS — no deletions.
• U3 no secrets: PASS — test-only diff contains no secret-shaped additions.
• U4 AC coverage: PASS — drives claim_review_task through kanban_reject and asserts ready recovery plus rejected run/event semantics.
• U5 mergeable: PASS — CLEAN.
• U6 UI verification: N/A — test-only Python change.
• C1 CI green: PASS — required CI and all eight Python test slices passed.
• C2 type-discipline: PASS — no new escapes in the diff.
• C3 lint clean: PASS — CI ruff/ty diff passed; independent ruff passed.
• C4 tests touched: PASS — test-only regression addition.
• C5 worker identity: PASS — commit 4f30f443 author and committer are Sahil (AI) <266772320+sahilm-ai@users.noreply.github.com>.

Code-quality judgment (role-reviewer)

Violations: None.

Independent evidence: scripts/run_tests.sh tests/tools/test_kanban_tools.py -q → 111 passed; ruff check tests/tools/test_kanban_tools.py → passed. Reviewed the changed test and its worker_env fixture, plus claim_review_task, reject_task, and _handle_reject. No Cursor/Bugbot inline findings were present.

---

Findings are mechanical (matrix) or judgment-based (role-reviewer). If a finding looks wrong, leave a counter-comment on the kanban task and Sahil will adjudicate on human-review.