- Add SkillSelectionResolver and SkillSelectionContext for deterministic skill sets
- Implement skill-resolver.ts with request/filter/resolve workflow
- Add session-skill-context.ts helper for unified skill selection across all session types
- Wire skill override into createKbAgent via AgentOptions
- Apply skill selection to triage, executor, reviewer, merger, and heartbeat sessions
- Add comprehensive tests for skill-resolver and session-skill-context
- Export skill resolver types and functions from engine index
- Add skill selection resolver pattern to memory documentation
An AI review agent (FN-1506) killed the running dashboard by finding
the process on port 4040 via lsof and running kill -9, causing exit
code 137 (SIGKILL) with no logs. This adds multi-layer guardrails:
- AGENTS.md: project-level rule reserving port 4040
- Executor/reviewer system prompts: explicit prohibition on killing
port 4040 processes, with instruction to use --port 0 instead
- Core agent-prompts.ts: same guardrails in all prompt variants
- Reviewer told to issue REVISE if executor violates the rule
- SIGHUP handlers in dashboard.ts and serve.ts for resilience
- Background engine reconciliation in dashboard/serve startup
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Background stale spec checks now skip tasks that are already in an
active work state (in-progress, in-review, done, merging, merging-pr)
to avoid interrupting work that is underway or complete.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add spec-staleness evaluator to check task specification age before execution
- Guard executor startup and resume to prevent running tasks with stale specs
- Guard scheduler dispatch to skip stale tasks and move them back to triage
- Add comprehensive tests for spec staleness detection (7d/14d thresholds)
- Add getFusionDir mock for executor staleness check tests
- Replace blocking execSync calls with async execAsync in executor.ts for user-configured commands
- Convert self-healing.ts worktree status checks to async to avoid blocking the event loop
- Update step-session-executor.ts to use async worktree operations
- Refactor worktree-pool.ts for fully async worktree creation, cleanup, and listing
- Convert pi.ts agent session handling to async execution
- Update all corresponding tests with async/await patterns
- Add changeset for @gsxdsm/fusion patch release
- Add context-overflow detection and recovery in step-session executor with same parity as single-session
- Implement step resumption from overflow checkpoints with accumulated context replay
- Add overflow state tracking and recovery logging for diagnostics
- Harden single-session executor overflow recovery with improved state management
- Update appendAgentLog parameter signatures across executors
- Add comprehensive tests for context overflow scenarios in both execution modes
- Add memory entry documenting the unified context-limit recovery approach
The merger's test/build verification, executor's worktreeInitCommand,
setupScript, and script-mode workflow steps all used execSync, which
blocks Node's event loop until the child process exits. A hanging
pnpm test could freeze the entire engine (no logs, heartbeats, or
other task progress) for the full 5-minute timeout.
Switch these call sites to promisify(exec) with awaited calls so the
engine keeps running while user-configured commands execute. Short
internal git plumbing (rev-parse, branch -d, worktree remove) still
uses execSync since those commands are bounded and measured in ms.
Document the rule in AGENTS.md under a new "Engine process rules"
section so future agents don't reintroduce blocking behavior.
Tests: update child_process mocks in merger.test, executor.test, and
restart.integration.test to route the new async exec through the
existing execSync mock and expose promisify.custom so destructuring
{ stdout, stderr } matches real child_process.exec semantics.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit adds the schema migration and types for the mission execution loop validation system:
- Adds loop state tracking columns to mission_features table (loopState, implementationAttemptCount, validatorAttemptCount, lastValidatorRunId, lastValidatorStatus, generatedFromFeatureId, generatedFromRunId)
- Creates mission_validator_runs table for tracking validation runs
- Creates mission_validator_failures table for assertion failure records
- Creates mission_fix_feature_lineage table for tracking fix feature relationships
- Adds workflowStepRetries column to tasks table for retry tracking
- Adds FEATURE_LOOP_STATES and VALIDATOR_RUN_STATUSES enums
- Updates TaskStore to support workflowStepRetries field
- Updates TaskExecutor to handle workflow step failures with retry logic
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
- Add ESLint configuration (eslint.config.mjs) for TypeScript/JavaScript linting
- Update executor prompts to include lint instruction before code submission
- Add lint check to triage prompt validation workflow
- Update agent prompts to emphasize lint compliance as quality requirement
- Add lint tool to agent toolset with file-level rule disabling capability
- Include lint in CI workflow with non-blocking status
- Update tests to verify lint-inclusive prompt behavior
- Add documentation for lint integration in contributing.md
- Add changeset for @gsxdsm/fusion minor release
- Add return statement after successful context compaction and resume
- When compact-and-resume succeeds (promptWithFallback completes without error),
executor now returns early instead of falling through to failure path
- This allows the finally block to clean up without marking the task as failed
- Add logging for recovery success and failure paths
- Add regression tests for context limit error detection
- Pattern is conservative: requires both 'context window' and 'exceeds' present
- Add shared RunAuditService in packages/engine/src/run-audit.ts for tracking agent run state transitions
- Instrument HeartbeatMonitor to log run mutations (create, complete, error, skip) with timestamps and context
- Instrument TaskExecutor to log run mutations during task execution lifecycle
- Instrument Merger to log run mutations during merge workflow
- Add run-audit pattern documentation to project memory
- Fix: remove audit calls from non-run recovery methods to avoid false positives
- Add assigneeUserId field to Task type and SQLite schema for human assignment
- Add reviewHandoffPolicy setting to control automatic handoff behavior
- Implement handoff detection in executor: detect user assignment during review and auto-transition task
- Add dashboard API routes for user assignment, handoff queries, and completion
- Add frontend API functions: getHandoffTask, assignTaskToUser, completeHandoff
- Add comprehensive tests for store methods, API routes, and executor handoff logic
- Update memory documentation with review handoff pattern
When a stuck task's session is disposed but an in-flight bash subprocess
keeps running, the executor's Promise never resolves, so the requeue
finally block never fires and the task stays stranded in-progress
indefinitely (observed with FN-1269).
Add a 60s safety-net timer in markStuckAborted: if the task is still in
executing after the grace period, force-requeue it directly by clearing
the worktree, setting status=stuck-killed, moving to todo, and removing
it from the executing set so the scheduler can re-dispatch normally.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add explicit always-green test suite instructions to executor agent prompts
- Update executor to enforce test-suite validation before task completion
- Add tests for agent prompt generation and executor behavior
- Ensure test failures block merge-ready state rather than allowing broken builds
store.logEntry, assignTask, and checkoutTask now receive an extra
runContext argument. Updated test assertions in executor.test.ts,
agent-heartbeat.test.ts, and restart.integration.test.ts to match
the current call signatures using expect.objectContaining where
appropriate.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add RunMutationContext type to track which agent run caused a mutation
- Thread runContext through TaskStore.logEntry, addComment, addSteeringComment, and pauseTask
- Propagate runContext from HeartbeatMonitor.executeHeartbeat to task store operations
- Propagate runContext from TaskExecutor.execute to task store operations
- Add GET /api/agents/:id/runs/:runId/mutations endpoint to query mutations by runId
- Add createTaskLogToolWithContext for heartbeat tools with run context support
- Add comprehensive tests for RunMutationContext across store and heartbeat modules
- Update memory.md with RunMutationContext usage convention
- Add Send Back dropdown to TaskCard showing for in-progress tasks with available target columns
- Thread onMoveTask prop through Column component to BoardView
- Terminate agent sessions when tasks move away from in-progress column
- Add executor tests for move-away session termination
- Add TaskCard send-back UI tests with dropdown visibility and interaction verification
- Add /missions/health endpoint handling to MissionManager test mocks
- Add listMissionsWithSummaries to mission-e2e test mock
- Add planState to Slice type and mock factories
- Add stuckKillCount to retry task test assertions
- Update log message for stuck-killed retry
The store's updateTask treats `undefined` as "no change" and `null` as
"clear field". Six places in executor.ts passed `undefined` when intending
to clear worktree/branch, leaving stale references that caused tasks to
get stuck in in-progress with no active session after stuck-kills, pauses,
and transient error retries.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
When context overflow or compaction causes an agent to lose awareness of
the task_done tool, the executor now checks if all steps are complete
before failing — treating it as an implicit task_done. Also adds
self-healing recovery for tasks that already slipped through as
misclassified failures in in-review.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Update executor, triage, and merger prompt strings to identify the product as "fn"
- Align branch/worktree documentation examples with fusion/fn-* naming conventions
- Rename dashboard websocket attachment guard from __kbWebSocketsAttached to __fnWebSocketsAttached
- Refresh engine tests, dashboard utility comments, and TUI header text to remove stale kb wording
- Wire AgentLogger into heartbeat runs, including tool/text callbacks and stdout excerpts on run records
- Persist heartbeat context taskId snapshots earlier and ensure logger flushes on success and failure paths
- Pass TaskStore into StepSessionExecutor and flush per-attempt agent logs in finally cleanup
- Expand heartbeat and step-session tests to verify log persistence and flush behavior, and add a patch changeset for @gsxdsm/fusion
- Add task_document_write and task_document_read factories with schemas, revision-aware responses, and error handling in agent-tools
- Wire document tools into executor and triage sessions and update prompts to persist and reuse planning artifacts across runs
- Export document tool factories and parameter schemas from engine public entrypoints while keeping executor re-export compatibility
- Add comprehensive tests covering write/read success paths, empty/not-found cases, and store failure handling
- Wrap TaskExecutor task:updated async listener logic in a top-level guard and log uncaught listener failures
- Add explicit catch handling for ProjectManager activity logging and dashboard stuck-detector settings-triggered checkNow calls
- Document async EventEmitter guard conventions in scheduler and hybrid executor listener wiring
- Add regression tests for executor and dashboard listener guards and include a patch changeset for @gsxdsm/fusion
- Move single-session and step-session executor failure paths to in-review after marking tasks failed
- Route exhausted transient recovery retries to in-review instead of leaving tasks outside review flow
- Move stuck-kill budget exhaustion failures in self-healing to in-review and update failure log wording
- Add regression tests in executor and self-healing suites to verify in-review transitions on these failure states
- Extend agent instruction resolution to append a Performance Feedback section with average score, trend, category breakdown, and recent comments when ratings exist
- Add rating-aware instruction resolution with graceful fallback when no store is configured, agent IDs are missing, or rating lookup fails
- Wire executor custom instruction loading to fetch agent rating summaries and include them in resolved instructions
- Expand agent-instructions tests to cover feedback formatting, trend indicators, comment limits, and fallback/error paths
- Extend project settings with reflection enablement, interval, and post-task trigger defaults
- Implement AgentReflectionService to gather agent/task history, generate structured AI reflections, and persist reflection metrics
- Add reflect_on_performance tool factory with optional focus area input and human-readable reflection output
- Wire reflection tool into TaskExecutor only when reflection is enabled and the task has an assigned agent
- Add comprehensive reflection service/tool tests and stabilize agent-store org tree expectation ordering
- Show resolved default executor, validator, and planning models in the task Model tab
- Pass merged settings into the Model tab from TaskDetailModal so default badges reflect effective runtime resolution
- Hot-swap active single-session executor models when task overrides change, including fallback to project defaults when overrides are cleared
- Add dashboard and executor tests for default model display, hot-swap behavior, fallback logic, and failure logging
- Document hot-swap behavior and limitations in AGENTS.md and add a patch changeset for @gsxdsm/fusion
- Add instructionsPath and instructionsText fields to Agent type and AgentStore
- Create agent-instructions resolver module in engine with priority-based resolution
- Wire custom instructions into executor, triage, reviewer, and merger agents
- Add PATCH /agents/:id/instructions API endpoint with file and text support
- Add instructions editor UI to dashboard agent detail config tab
- Add comprehensive tests for instructions resolver and AgentStore integration
- Add changeset for published package bump
- Extract per-task thinkingLevel from task detail, falling back to global defaultThinkingLevel
- Apply override in initial execution, retry session, and spawned agent paths
- Add 3 tests: per-task override, global fallback, explicit 'off' overrides global
- Fix step status not being tracked correctly in executor callbacks (done/skipped were not updated)
- Expose runStepsInNewSessions and maxParallelSteps settings in CLI settings command
- Add AGENTS.md documentation for runStepsInNewSessions and maxParallelSteps settings
- Add changeset for the published CLI package
- Add tests for CLI settings validation and executor step status tracking
- Extract shared agent tools (bash, read, write, edit) from executor into agent-tools.ts for reuse
- Implement heartbeat execution in HeartbeatMonitor with configurable interval and task_done support
- Wire heartbeat execution into InProcessRuntime lifecycle (start/stop)
- Add comprehensive test suite covering execution, timeouts, concurrency, and error handling
- Add heartbeat-related log constants to logger
- Add AgentPromptTemplate and AgentPromptsConfig types to ProjectSettings
- Create agent-prompts module with 7 built-in prompt templates and role resolver
- Wire engine agents (executor, reviewer, merger, triage) to use resolved prompts
- Add 25 test cases covering template resolution, role assignment, and validation
- Export new types from @fusion/core package
- Document agentPrompts configuration and built-in templates in AGENTS.md
