|
289 | 289 | {"id":"bd-krew","title":"doctor: zombie run detection is broken for SDK-based agent workers","description":"In src/orchestrator/doctor.ts, extractPid() uses regex /pid-(\\d+)/ to extract a PID from the session_key. However, SDK-based agent workers use session_key format 'foreman:sdk:<model>:<uuid>' which contains no PID. extractPid() returns null for all SDK runs, so isProcessAlive() is called with null which makes isAlive=false, causing ALL SDK-based running runs to be reported as zombies — even when live agent-worker.ts processes are actively running them. The fix must detect liveness differently for SDK runners, e.g. by checking for a running agent-worker.ts process whose worker JSON file contains the run ID, or by storing the actual PID in a separate column.","status":"closed","priority":0,"issue_type":"bug","created_at":"2026-03-18T03:58:33.651572Z","created_by":"ldangelo","updated_at":"2026-03-20T04:42:32.293139Z","closed_at":"2026-03-20T04:42:32.291436Z","close_reason":"Completed via pipeline","source_repo":".","compaction_level":0,"original_size":0} |
290 | 290 | {"id":"bd-krv","title":"[trd:seeds-to-br-bv-migration:task:TRD-012-TEST] Unit tests for dispatcher prompt content","description":"## Test Task: TRD-012-TEST\nTRD Reference: docs/TRD/seeds-to-br-bv-migration.md#trd-012-test\nPRD Reference: docs/PRD/PRD-2026-001-seeds-to-br-bv-migration.md#req-016\nVerifies: TRD-012\nSatisfies: REQ-016\nTarget Files: src/orchestrator/__tests__/dispatcher.test.ts\nActions:\n1. Test spawnAgent prompt contains \"br close\" when backend=br\n2. Test resumeAgent prompt contains \"br close\" when backend=br\n3. Test no \"sd close\" in prompts when backend=br\nDependencies: TRD-012","status":"closed","priority":2,"issue_type":"task","created_at":"2026-03-16T13:23:25.285445Z","created_by":"ldangelo","updated_at":"2026-03-16T17:11:15.750338Z","closed_at":"2026-03-16T17:11:15.750015Z","close_reason":"Tests implemented and passing — 12 tests in dispatcher-prompts.test.ts","source_repo":".","compaction_level":0,"original_size":0,"dependencies":[{"issue_id":"bd-krv","depends_on_id":"bd-18m","type":"blocks","created_at":"2026-03-16T13:23:25.564508Z","created_by":"ldangelo","metadata":"{}","thread_id":""}]} |
291 | 291 | {"id":"bd-ksbk","title":"doctor: checkFailedStuckRuns has no --fix path; 155 failed runs accumulate indefinitely","description":"checkFailedStuckRuns() in doctor.ts:496 only warns about failed/stuck runs but has no fix logic — it takes no opts parameter and never cleans anything up. Running 'foreman doctor --fix' leaves the warning unchanged.\n\nAfter the bd-qtqs retry loop bug (bd-zwtr), 155 failed runs accumulated in the DB. doctor --fix cannot remove them. They persist indefinitely and noise the doctor output.\n\nTwo fixes needed:\n1. Add fix option to checkFailedStuckRuns() to purge/archive failed runs older than N days (e.g. 7 days) for seeds that are now closed or have a successful later run.\n2. Or: filter getRunsByStatus('failed') to exclude runs where the seed has a subsequent 'completed' or 'merged' run — these are historical retries, not actionable failures.\n\nThe check should also distinguish between:\n- Seeds with ONLY failed runs (actionable — needs attention)\n- Seeds with failed runs BUT ALSO a later completed/merged run (noise — historical retries)","notes":"[FAILED] [DEVELOPER] ","status":"closed","priority":2,"issue_type":"bug","created_at":"2026-03-19T15:27:26.999025Z","created_by":"ldangelo","updated_at":"2026-03-23T20:12:01.296252Z","closed_at":"2026-03-23T20:12:01.295511Z","close_reason":"done","source_repo":".","compaction_level":0,"original_size":0} |
292 | | -{"id":"bd-kx19","title":"Test: Verify release workflow with dry-run","description":"Run the release workflow in dry-run mode. Verify version detection from commit history. Verify npm pack produces correct package. Verify binary build matrix produces all 5 targets. Test on a non-main branch first.","notes":"Post-merge tests failed (attempt 2/3). Will retry after the developer addresses the failures. \nFirst failure:\n\n> @oftheangels/foreman@0.1.0 test\n> vitest run\n\n\n\u001b[1m\u001b[46m RUN \u001b[49m\u001b[22m \u001b[36mv4.0.18 \u001b[39m\u001b[90m/Users/ldangelo/Development/Fortium/foreman\u001b[39m\n\n \u001b[31m❯\u001b[39m src/orchestrator/__tests__/task-backend-ops.test.ts \u001b[2m(\u001b[22m\u001b[2m48 tests\u001b[22m\u001b[2m | \u001b[22m\u001b[31m1 failed\u001b[39m\u001b[2m)\u001b[22m\u001b[32m 14\u001b[2mms\u001b[22m\u001b[39m\n \u001b[32m✓\u001b[39m calls br close --no-db --force with seedId (beads_rust#204 workaround)\u001b[32m 2\u001b[2mms\u001b[22m\u001b[39m\n \u001b[32m✓\u001b[39m uses ~/.local/bin/br path for br backend\u001b[32m 0\u001b[2mms\u001b[22m\u001b[39m\n ","status":"blocked","priority":2,"issue_type":"task","created_at":"2026-03-24T02:29:32.471349Z","created_by":"ldangelo","updated_at":"2026-03-25T12:34:30.617502Z","source_repo":".","compaction_level":0,"original_size":0,"labels":["phase:developer","phase:finalize","phase:qa","phase:reviewer"],"dependencies":[{"issue_id":"bd-kx19","depends_on_id":"bd-i3c9","type":"blocks","created_at":"2026-03-24T02:29:50.791655Z","created_by":"ldangelo","metadata":"{}","thread_id":""}]} |
| 292 | +{"id":"bd-kx19","title":"Test: Verify release workflow with dry-run","description":"Run the release workflow in dry-run mode. Verify version detection from commit history. Verify npm pack produces correct package. Verify binary build matrix produces all 5 targets. Test on a non-main branch first.","notes":"Post-merge tests failed (attempt 2/3). Will retry after the developer addresses the failures. \nFirst failure:\n\n> @oftheangels/foreman@0.1.0 test\n> vitest run\n\n\n\u001b[1m\u001b[46m RUN \u001b[49m\u001b[22m \u001b[36mv4.0.18 \u001b[39m\u001b[90m/Users/ldangelo/Development/Fortium/foreman\u001b[39m\n\n \u001b[31m❯\u001b[39m src/orchestrator/__tests__/task-backend-ops.test.ts \u001b[2m(\u001b[22m\u001b[2m48 tests\u001b[22m\u001b[2m | \u001b[22m\u001b[31m1 failed\u001b[39m\u001b[2m)\u001b[22m\u001b[32m 14\u001b[2mms\u001b[22m\u001b[39m\n \u001b[32m✓\u001b[39m calls br close --no-db --force with seedId (beads_rust#204 workaround)\u001b[32m 2\u001b[2mms\u001b[22m\u001b[39m\n \u001b[32m✓\u001b[39m uses ~/.local/bin/br path for br backend\u001b[32m 0\u001b[2mms\u001b[22m\u001b[39m\n ","status":"closed","priority":2,"issue_type":"task","created_at":"2026-03-24T02:29:32.471349Z","created_by":"ldangelo","updated_at":"2026-03-25T13:33:32.258273Z","closed_at":"2026-03-25T13:33:32.257815Z","source_repo":".","compaction_level":0,"original_size":0,"labels":["phase:developer","phase:explorer","phase:finalize","phase:qa","phase:reviewer"]} |
293 | 293 | {"id":"bd-l0w","title":"Rename 'seed' terminology to 'bead' throughout codebase","description":"The codebase inconsistently uses 'seed' and 'bead' to refer to the same concept (a br/beads task ID). The user-facing CLI uses --seed flags and 'seed_id' internally, but the underlying task tracker is called 'beads' (br). This creates confusing UX and inconsistent naming.\n\n## Scope of Changes\n\n### CLI flags\n- foreman run --seed <id> → foreman run --bead <id>\n- foreman merge --seed <id> → foreman merge --bead <id>\n- foreman reset --seed <id> → foreman reset --bead <id>\n- foreman status --seed <id> → foreman status --bead <id>\n- Any other --seed flags in CLI commands\n\n### Internal naming (TypeScript)\n- seed_id column references in SQLite queries → bead_id (or rename column via migration)\n- Run.seed_id field → Run.bead_id\n- opts.seed → opts.bead\n- seedId variables → beadId\n- getRunsByStatus / getRunsBySeed → update param names\n- MergeQueueEntry.seed_id → bead_id\n- mergeCompleted({ seedId }) → mergeCompleted({ beadId })\n- getCompletedRuns(projectId, seedId) → getCompletedRuns(projectId, beadId)\n- resetForRetry(seedId) → resetForRetry(beadId)\n- All other internal seed_id / seedId references\n\n### Worker config / agent-worker\n- WorkerConfig.seed_id → WorkerConfig.bead_id\n- All references in dispatcher.ts, agent-worker.ts, monitor.ts\n\n### Store schema migration\n- Add SQLite migration: ALTER TABLE runs RENAME COLUMN seed_id TO bead_id\n- Add SQLite migration: ALTER TABLE merge_queue RENAME COLUMN seed_id TO bead_id\n- Ensure migration runs on first open of existing databases\n\n### Tests\n- Update all test fixtures, mocks, and assertions that reference seed_id / seedId / opts.seed\n\n### Backwards compatibility\n- Accept --seed as a deprecated alias for --bead with a deprecation warning during a transition period\n\n## Acceptance Criteria\n- foreman run --bead <id> dispatches the correct task\n- foreman merge --bead <id> merges the correct branch\n- Internal code uses beadId/bead_id consistently\n- All existing tests pass with updated naming\n- --seed still works but prints deprecation warning","status":"closed","priority":2,"issue_type":"chore","created_at":"2026-03-17T18:01:59.244732Z","created_by":"ldangelo","updated_at":"2026-03-17T18:35:07.961623Z","closed_at":"2026-03-17T18:35:07.961252Z","close_reason":"Completed via pipeline","source_repo":".","compaction_level":0,"original_size":0} |
294 | 294 | {"id":"bd-l5r9","title":"[trd:trd-2026-002-pi-agent-mail-rpc-migration:task:TRD-017] Pi Extension Health Check","description":"TRD Reference: docs/TRD/TRD-2026-002-pi-agent-mail-rpc-migration.md#trd-017\\nSatisfies: REQ-018\\nValidates PRD ACs: AC-018-3\\nTarget File: src/orchestrator/pi-rpc-spawn-strategy.ts\\nActions:\\n1. Send health check RPC command after Pi session initialized\\n2. Verify foreman-tool-gate is in loaded extension list\\n3. Refuse to start pipeline and log actionable error if not loaded\\nDependencies: TRD-012\\nEst: 2h","status":"closed","priority":3,"issue_type":"task","created_at":"2026-03-19T23:52:37.310074Z","created_by":"ldangelo","updated_at":"2026-03-20T02:34:59.172087Z","closed_at":"2026-03-20T02:34:59.171628Z","close_reason":"done","source_repo":".","compaction_level":0,"original_size":0,"dependencies":[{"issue_id":"bd-l5r9","depends_on_id":"bd-kkw0","type":"blocks","created_at":"2026-03-19T23:53:44.352120Z","created_by":"ldangelo","metadata":"{}","thread_id":""}]} |
295 | 295 | {"id":"bd-l72","title":"Move bead lifecycle ownership to agent-worker — eliminate dispatcher/worker split","description":"Currently bead status management is split: dispatcher.ts marks in_progress at line 186, agent-worker.ts finalizes via br close, and reset.ts resets to open. This causes race conditions (bd-ng9) and missing updates (bd-7wa). Consolidate: 1) Remove seeds.update(in_progress) from dispatcher.ts — instead pass seeds/br client config to agent-worker via WorkerConfig, 2) agent-worker marks in_progress when starting (before explorer phase), 3) agent-worker resets to open on failure/stuck (currently it only updates SQLite), 4) agent-worker calls br close on success (already does this in finalize). This makes agent-worker the single owner of bead lifecycle, eliminating the race condition.","status":"closed","priority":1,"issue_type":"feature","created_at":"2026-03-17T21:31:18.293001Z","created_by":"ldangelo","updated_at":"2026-03-20T04:42:47.980916Z","closed_at":"2026-03-20T04:42:47.979525Z","close_reason":"Completed via pipeline","source_repo":".","compaction_level":0,"original_size":0} |
|
0 commit comments