## Summary
PostgreSQL runtime roles without `CREATE` permission on `public` no
longer trigger schema writes during migration-marker health reads, so
`permission denied for schema public` is not mislabeled as database
corruption. Once connectivity and task-ID integrity pass, an unavailable
migration marker is treated as advisory instead of making the whole
database unhealthy. Dashboard and notification guidance now describes a
PostgreSQL health failure accurately and renders actionable log and
recovery links in every supported locale.
## Validation
- 54 targeted tests passed across core, dashboard, engine, and i18n.
- Typechecks passed for all four affected packages.
- Scoped ESLint, strict changeset validation, and diff checks passed.
---
[](https://github.com/EveryInc/compound-engineering-plugin)
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* PostgreSQL health failures are now reported as degraded health checks
rather than database corruption.
* Migration-status lookup failures no longer incorrectly mark an
otherwise healthy database as unhealthy.
* Migration-state checks are now read-only and avoid creating or
modifying database structures.
* **UI & Localization**
* Updated database health banner messaging and recovery guidance across
supported languages.
* The banner now appears for broader PostgreSQL health failures and
links to storage documentation.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
## Summary
Pooled worktrees no longer retain the previous task's identity after
reassignment, preventing valid commits from being rejected when the
checked-out branch and stale ownership marker disagree. Fusion now
refreshes the identity guard immediately after the pool prepares the new
branch and before the checkout is exposed to the task, preserving the
cross-task commit safety check. Regression coverage exercises the pooled
acquisition path and confirms the new task identity is installed.
Related: FN-8400
## Validation
- 85 focused engine tests passed across executor worktree and
acquisition coverage.
- `@fusion/engine` typecheck passed.
- Changed implementation lint and changeset validation passed.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* Prevented stale pooled worktree ownership metadata from blocking
commits after a pooled checkout is reassigned.
* Refreshed task identity metadata when pooled worktrees are reused
across task branches, ensuring hooks/attribution settings are correctly
applied.
* **Tests**
* Added/updated coverage to confirm the task identity guard is
reinstalled when acquiring pooled worktrees (including a real-git
scenario proving commits succeed after stale identity is cleared).
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Part **2 of 3** of the IR-driven lifecycle cutover (stacked on #2341;
top is #2335).
**Scope (80 files, packages/engine + cli/pi skill docs +
AGENTS/architecture):** graph-driven column moves via the
column-boundary controller (R1), single-mover scheduler/hold-release
trait cutover (KTD-2/KTD-9), trait re-keyed self-healing + merger with
the R7b confirmed-merge-must-finalize guarantee, graph-exclusive Plan
Review with leased dedup (R4/R5), the executeCore body-lift — zero
legacy re-entry — with fn_review_step + interceptor machinery deleted
and tombstone-ratcheted (R9), builtin workflow runtime fixes (missing
hold handler, unseamed-node column inheritance, no-merge completion
mover), the 6-column benchmark acceptance suite (11 tests) + 12-builtin
lifecycle sweep (94 assertions), and the executor test-harness
modernization. Also retires core's interpreter-cutover scaffolding whose
last consumer (the authoritative driver) dies here.
**Merge order:** #2341 → this → #2335. After #2341 merges, retarget this
to main.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
---------
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
## Summary
Main Full Suite shards have been red after recent landings. Root causes:
1. **Executor tests** — `execute()` now polls
`getTaskVerificationRequestAsync` (chat-enqueued verification). Shared
`createMockStore()` (and soft-delete inline store) lacked the method, so
nearly every execute-path suite failed with `is not a function`.
2. **TaskDetailModal suites** — `NativeStructurePreview` imports `Map` /
`Lightbulb` / `BarChart3` / `Target` / `CircleAlert` from lucide; the
shared TaskDetail lucide mock omitted them, so suites failed at import.
3. **Grok process-lifecycle** — 15s bound stress timed out under
full-suite load without product-bug evidence → quarantined on sight per
AGENTS.md.
## Test plan
- [x] `executor-task-done-blocked`, `executor-fast-mode-workflows`,
concurrent-execute race
- [x] `executor-step-session`, plan-only scope leak, review-step
indexing
- [x] `TaskDetailModal.create-pr` + `TaskDetail.mobile-transition`
- [ ] Full Suite CI on this PR
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Improvements**
* Added `html2canvas` support in the dashboard to enable HTML-to-canvas
rendering needed for visual structure previews.
* **Tests**
* Updated task execution test mocks to handle task verification-request
flows reliably.
* Improved task deletion safeguard coverage and related execution
behavior checks.
* Enhanced test stubs to support structure preview rendering elements
during modal-related tests.
* **Chores**
* Quarantined a timing-sensitive process lifecycle test and refreshed
quarantine tracking to improve full-suite stability.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Retry MCP bootstrap with bounded backoff and request deadlines, then continue with healthy servers when an integration remains unavailable.
Exclude definitions with unresolved secrets so graceful degradation never connects a partially authenticated server.
Serialize cross-parent diagnostic claims at project scope and reuse a recent active canonical task. Keep ordinary, completed, and path-distinct work independent.
## Summary
Fixes a race where a stuck-kill during Plan Review could leave an
approved task stranded in `todo` as **unplanned**, so the scheduler
refused release to `in-progress` for several minutes (observed on
FN-1312: ~6m39s idle after Plan Review APPROVE).
### What went wrong
1. Finalize clears `status` early, then runs Plan Review.
2. Stuck-kill disposes the main triage session near the 30m processing
threshold.
3. Stale eviction only treated the main session as live, so the card
left `processing` while Plan Review / finalize was still running.
4. A second planner claimed `status: "planning"` and rewrote
`PROMPT.md`.
5. The first finalize moved `triage → todo` without clearing planning
statuses.
6. Hold-release saw planning/`needs-replan` and blocked: *“card is
unplanned and cannot enter processing column in-progress”*.
### Fix
In `packages/engine/src/triage.ts`:
- Track **finalizing** handoffs and **live Plan Review subagents** as
live planning work.
- Keep them in `getProcessingTaskIds`, refuse concurrent `specifyTask`,
and skip stale-processing eviction while they are live.
- **Defer** stuck-abort requeue during finalize (bump kill count only;
do not force `needs-replan`).
- Recover approved drafts with `status: null` (post early-clear), not
only `status: "planning"`.
- Re-assert `status: null` after the successful release move to todo.
### Tests
Regression coverage in `triage.test.ts` for eviction retention,
processing-id inclusion, null-status recovery, needs-replan
non-recovery, and deferred stuck-abort during finalize.
## Test plan
- [x] `pnpm --filter @fusion/engine exec vitest run
src/__tests__/triage.test.ts -t "evictStaleProcessing|specified triage
recovery|stuck|recoverApproved|markStuckAborted|defers stuck-abort"`
(related cases green)
- [ ] CI gate on this PR
- [ ] Optional: reproduce stuck-kill mid–Plan Review and confirm todo
releases without a second full replan
## Notes
Secondary report (stuck-kill immediately after all implementation steps
complete, then 16s auto-recovery) is **out of scope** here; existing
recovery already continued by skipping completed steps.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* Improved recovery for approved tasks when Plan Review has completed,
including correct handling of `null`-status drafts.
* Avoided recovering unplanned seed drafts and tasks that should enter a
replanning flow.
* Refined stuck-abort requeue/cleanup to respect in-progress Plan Review
and finalize handoffs.
* Prevented stale-processing eviction from removing tasks while Plan
Review/subagent finalization is still active.
* Blocked new triage discovery and task specification when finalize/Plan
Review work is ongoing, preventing handoff disruption.
* Preserved durable task state during handoff completion
(approval/failure/replan outcomes).
* **Tests**
* Added expanded triage recovery and stuck-abort/stale-eviction
regression coverage.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Prevent concurrent Fusion sessions and processes from consuming the same rotating refresh token. Share one refresh lock across legacy and subscription Anthropic credential aliases.
## Summary
- Heartbeat customTools inventory includes `fn_research_promote_finding`
(64 tools).
- Schema-applier bookkeeping for migration `0023` research feature
provenance.
- Seed upgrade fixtures include `mission_features.slice_id` for the
unique index.
## Evidence
- Full greens earlier this session: **29662309385** (#2323),
**29663526777** (#2325), **29664692957** (#2327).
- Tip red after FN-8297: heartbeat 63→64 + schema upgrade seed.
## Test plan
- [x] heartbeat expected-tools + schema upgrade tests green locally
- [ ] Full Suite all shards green on tip after merge
## Summary
Planning breakdowns now preserve their creating task as durable lineage
and reuse only siblings from that same parent. Identical wording under a
different parent creates a distinct child instead of silently linking
the wrong lineage.
The dashboard planning path now uses the same duplicate-safe creation
contract as agent tools, leaves reused canonical tasks untouched, and
exposes API-created parent links in task details.
Related: FN-8277
## Validation
- Core duplicate guard: 12 tests passed
- Engine task creation: 32 tests passed
- Dashboard planning routes: 4 focused tests passed
- Dashboard task detail provenance: 2 focused tests passed
- Core, engine, dashboard, and CLI typechecks passed
- Lint and strict changeset validation passed
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
- **New Features**
- Preserved parent-task lineage for subtasks created through planning
breakdowns and API workflows.
- Improved duplicate detection so identical tasks from different parent
tasks can coexist safely.
- Added parent-task links to API-created task provenance details.
- Reused existing duplicates only within the same parent-task context.
- **Bug Fixes**
- Prevented duplicate handling from incorrectly archiving or skipping
tasks belonging to other parents.
- Improved dependency handling when creating planned subtasks.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
## Summary
Moving an active task back to Todo could update the board before its
agent and subprocesses had stopped, leaving a Todo card that was still
processing. User-initiated in-progress-to-Todo moves now wait for every
executor cancellation surface before the new column is persisted or
returned to the dashboard. Cancellation is fail-closed and bounded: a
wedged shutdown leaves the task in Progress, releases its lock for
recovery, and fences late cleanup from replacement execution
generations. Engine-driven recovery moves and other transitions retain
their existing behavior.
## Validation
- Confirmed with PostgreSQL-backed symptom tests that the durable row
stays in Progress while cancellation is pending and that a timeout
releases the task lock without publishing Todo.
- Verified multi-executor ownership and replacement-generation fencing
across focused core and engine tests: 18 tests passed.
- Core build, engine typecheck, targeted lint, and strict changeset
validation passed.
---
[](https://github.com/EveryInc/compound-engineering-plugin)
## Summary
- FN-8277 parent-scoped uniqueness: mock
`findRecentTasksBySourceParentTaskId` in heartbeat/triage/split suites.
- FN-8326: index `reportRoadmapDedup` in Settings search.
- Quarantine re-flaked `dev-server-process` under full-suite API load
(run 29661202279).
## Test plan
- [x] Targeted createTask / search-index tests green locally
- [ ] Full Suite all 4 shards green on main after merge
## Summary
Repeated review and executor sessions could replay a follow-up creation
step and produce another live task whenever the wording changed. In the
incident behind this fix, 21 creation calls for three intended
follow-ups left 18 duplicate tasks.
Agent-created tasks now retain their parent and agent provenance across
step sessions, heartbeats, and the published CLI surface. Same-parent
paraphrases converge on the existing task through a serialized pre-check
and a database-backed intent claim, while distinct sibling actions
remain separate. Candidate lookup is parent-indexed, uniqueness failures
abort creation, and canonical reuse no longer emits misleading creation
audit events or workflow claims.
Related: FN-8277
## Validation
- Core duplicate guard and intake: 24 tests passed
- Engine task creation and heartbeat: 154 tests passed
- CLI extension: 66 tests passed, 95 skipped
- Core, engine, and CLI typechecks passed
- Core, engine, and CLI builds passed before rebase; the rebase was
conflict-free
- Scoped lint, strict changeset validation, and diff checks passed
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* Prevented retried agent steps from creating duplicate follow-up tasks.
* Improved parent-scoped deduplication for paraphrased follow-ups while
preserving distinct sibling actions.
* Preserved parent-task and agent context for created follow-ups.
* Concurrent follow-up requests are now serialized/deduplicated so
duplicates link to the existing task instead of showing as newly
created.
* Updated agent follow-up/heartbeat activity so reused follow-ups no
longer appear in run results as fresh creations.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
## Summary
- Heartbeat customTools inventory includes FN-8295 ideation tools (63
total).
- Non-en i18n parity for FN-8286 `reviewArtifacts` Command Center +
settings keys.
- Quarantine `TaskDetailModal.tab-persistence.test.tsx` (CI load flake;
green focused thrice).
## Test plan
- [x] heartbeat expected-tools case green
- [x] i18n-gate-coverage + parity green
- [ ] Full Suite all 4 shards green on main after merge
## Summary
- Align heartbeat `customTools` expectations with FN-8294 mission
hierarchy tools (43→58).
- Refresh `COORDINATION_EXEMPT_TOOLS` snapshot for `fn_mission_list` /
`fn_mission_show`.
- Backfill `commandCenter.portability.*` for non-en locales and map
`reportMode` / `reportModeByAction` / `embeddedPostgresMaxConnections`
into settings default-description inventory with i18n help text.
- Realign FN-8064 skip-narration unit test with store-owned proactive
chat (no tool-side `appendAgentLog`).
- Quarantine load-sensitive `async-quality-store.pg.test.ts` (5s timeout
+ leftover psql under full-suite shard load; run 29657633544).
## Test plan
- [x] `pnpm --filter @fusion/engine exec vitest run`
gating-classifications + executor-prompt + heartbeat expected-tools case
- [x] `pnpm --filter @fusion/i18n exec vitest run` i18n-gate-coverage +
parity
- [x] `pnpm --filter @fusion/dashboard exec vitest run`
settings-default-descriptions
- [ ] Full Suite all 4 shards green on main after merge
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Settings**
* Added clearer, localized help text for report modes and per-action
overrides, including inheritance behavior.
* Added advanced embedded database connection-limit settings and
validation guidance.
* **Localization**
* Expanded translations for report settings, database tuning, and
organization configuration import/export workflows across supported
languages.
* **Tests & Maintenance**
* Updated test coverage and expectations for expanded tools and
reporting behavior.
* Quarantined a flaky database-related test.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
## Summary
Follow-up to #2275 review feedback: `getTask` mocks for DUPLICATE
recovery tests now return a default current-task fixture for
non-canonical IDs instead of `undefined`.
## Context
CodeRabbit on #2275 noted that returning `undefined` for every id except
the canonical task could break recovery if the subject task is
re-fetched mid-path.
## Test plan
- [x] `triage-split-into-subtasks-delete.test.ts`
- [x] `triage-finalize-duplicate-lineage.test.ts`
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Tests**
* Improved duplicate-resolution test coverage by preserving task data
during recovery and re-fetch scenarios.
* Added validation for non-canonical task handling to prevent errors
when duplicate tasks are resolved or split.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
