Adds the agentation floating annotation toolbar so annotated UI elements
produce structured markdown (CSS selectors, positions, React component info)
that can be copied/pasted into Claude to provide precise visual feedback.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add BUILDING, DISPATCHING, EXECUTING, HANDLING, LOGGING, EXHAUSTED_IDLE states
- These represent the inner loop of bead execution, all map to WORKING display
- DirectoryTailer now re-reads files modified within 4 hours from start on startup
This reconstructs worker state after FABRIC restart without replaying ancient history
- Update VALID_TRANSITIONS to include new state transitions
- Update color/icon mappings for new states
The test 'resumes from saved position when a file is re-activated after
eviction' was incorrectly creating both files before starting the tailer.
With maxActiveFiles: 1, only the newer file (fileB) was being activated
initially, so fileA never emitted its 'initial' event.
Restructured to:
1. Create fileA with content
2. Start tailer (fileA gets activated)
3. Wait for fileA to emit 'initial'
4. Create fileB (triggers eviction of fileA via dirWatcher)
5. Continue with re-activation test
This properly tests the LRU eviction and position checkpointing behavior.
The getEventTime function had a bug where it recursively called itself
when event.timestamp was truthy, causing "Maximum call stack size exceeded".
Fixed by using Date.parse() to convert the ISO timestamp string to unix ms.
All 26 failing TimelineView tests now pass.
Closes: bf-50m5
Implements the missing /api/spans/dag endpoint that was blocking the
SpanDag component. The endpoint queries span events from the store and
builds a hierarchical tree structure for visualization.
Changes:
- Added GET /api/spans/dag endpoint in src/web/server.ts
- Added SpanDagResponse interface to src/types.ts for JSON serialization
- Updated SpanNode interface to use nullable fields (null instead of undefined)
- Fixed src/dagUtils.ts to use nullable SpanNode fields
The endpoint accepts an optional trace_id query parameter to filter
spans by trace, and returns a SpanDagResponse with root spans, total
span count, and trace summary.
Closes: bf-82u8
The test adds 10001 events to verify the default maxEvents limit of 10000.
Each event triggers significant processing (collision detection, file
tracking, multiple manager updates), so the default 5s timeout was too
short. Increased to 30s; actual runtime is ~750ms.
Fixes timeout failure in src/store.test.ts > InMemoryEventStore > maxEvents limit
The test 'should use default maxEvents of 10000' was timing out when run
with the full test suite but passed in isolation. Root cause: global
singleton instances (WorkerAnalytics, CrossReferenceManager, etc.)
retained state across tests in the main 'InMemoryEventStore' describe
block.
Added afterEach hook that calls all available reset* functions to
ensure clean state between tests.
Closes: bf-5u6j
Phase 8 (Post-launch Fixes, bd-0nd series) was already complete but the
status line only mentioned Phases 1-7 and 9. Updated to show all phases
complete.
Backend API endpoints (src/web/server.ts):
- GET /api/workers/compare?worker1=&worker2= — returns WorkerComparison via analyticsManager.compareWorkers()
- GET /api/analytics/workers — returns per-worker WorkerMetrics for leaderboard table
- GET /api/analytics/sessions — exposes historicalStore.getSessions() for cross-session comparisons
Frontend component (src/web/frontend/src/components/WorkerAnalyticsPanel.tsx):
- Comparison view mirroring TUI WorkerAnalyticsPanel behavior
- Leaderboard table with sortable columns
- Historical sessions list
- Worker selection for comparison with diff/percent/winner indicators
Wired into App.tsx with new "Workers" button (⚔️ icon) and command palette action (show:worker-analytics)
Closes: bf-4cqq
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Implements Phase 6 Historical session index for comparisons.
Backend (src/web/server.ts):
- GET /api/sessions — list sessions (paginated, with start/end filter)
- GET /api/sessions/:id — get single session detail with per-worker summaries
- GET /api/sessions/:id/workers — get worker summaries for a session
All endpoints use the existing HistoricalStore infrastructure.
Frontend (src/web/frontend/src/components/HistoricalSessionsPanel.tsx):
- Sessions list table with duration, workers, tasks, cost, tokens, time range
- Click-to-expand session detail with worker performance breakdown
- Metrics source badge (otlp-metric, otlp-span, log-derived)
- Empty state with helpful hint when no sessions exist
- Refresh button for manual reload
Integration:
- Added to App.tsx with Sessions toggle button in header
- Command palette action: show:sessions
- Follows existing panel patterns (ProductivityPanel, AnalyticsDashboard)
Closes: bf-5xch
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
- Add currentBead: null to all WorkerInfo test fixtures (8 files)
- Add missing required fields to SemanticNarrative test fixtures
- accomplishments, challenges, sentiment, stats, generatedAt, isLive
- Add missing workerId and events to NarrativeSegment fixtures
- Fix onSelectCallback mock type assertion
- Add Record<string, string> index signature to mockBeadsData
All npx tsc --noEmit errors resolved. Test failures (6) remain
and are tracked in separate bead bf-40cu.
Closes: bf-7x4z
- Removed empty bin/fabric file (not used; package.json bin declaration is correct)
- Updated fabric-web.service and fabric-prune.service to use /home/coding/.nix-profile/bin/node instead of /usr/bin/node (NixOS node path)
- Created ~/.config/fabric/secrets.env with FABRIC_AUTH_TOKEN
- Installed and enabled fabric-web.service and fabric-prune.timer
Acceptance verified:
- systemctl --user status fabric-web.service shows active (running)
- curl http://localhost:3000/api/workers returns valid JSON ([])
Closes: bf-1nah
All Phase 9 items verified as implemented:
- beadsCompleted fires on bead.released/release_success
- currentBead field tracks active bead per worker
- Fleet summary bar shows real-time fleet state
- Worker cards show beadsCompleted + currentBead (removed eventCount)
- Worker sort by state (WORKING > SELECTING > EXHAUSTED)
- Test worker filter with hideTestWorkers toggle
- Productivity panel with daily throughput chart + worker leaderboard
- Bead workspace scanner reads .beads/issues.jsonl for project breakdown
- GET /api/productivity endpoint returns all productivity data
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
All remaining items verified complete:
- currentBead field (tracked in store.ts, displayed in WorkerGrid)
- Fleet summary bar (FleetSummaryBar.tsx, integrated in App.tsx)
- Worker card enrichment (beadsCompleted + currentBead shown)
- Bead workspace scanner (scanBeadWorkspaces for project breakdown)
All Phase 9 items are now fully implemented and functional.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Bead-Id: bf-2wf
Use vi.hoisted() to declare mock classes before the factory runs,
preventing "Cannot access X before initialization" errors that occur
when vitest hoists vi.mock() calls but the factory captures references
to variables not yet initialized.
Fixes errors in:
- src/tui/components/CrossReferencePanel.test.ts (MockCrossReferenceManager)
- src/tui/components/WorkerAnalyticsPanel.test.ts (MockWorkerAnalytics)
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Verified that the vitest mock hoisting fix remains in place:
- CrossReferencePanel.test.ts: 43 tests pass
- WorkerAnalyticsPanel.test.ts: 64 tests pass
- Total: 107 tests pass
The inline mock definitions in vi.mock() factories prevent
hoisting issues as documented in the note.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Updated note to reflect that the vitest mock hoisting fix is in place
using inline mock definitions (commit 55df248). No code changes required.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Fix mock hoisting issues in CrossReferencePanel and WorkerAnalyticsPanel test files by:
- Defining mock classes entirely inline within vi.mock() factory functions
- Removing references to external consts that were causing "Cannot access before initialization" errors
- Simplifying getMockFunctions() helper to use the already-imported mocked constructor
The vitest hoisting mechanism moves vi.mock() calls to the top of the file before regular variable declarations, so any consts referenced in the mock factory must also be defined inside the factory or via vi.hoisted(). The simplest fix is to define everything inline.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Fix "Cannot access before initialization" errors in:
- CrossReferencePanel.test.ts
- WorkerAnalyticsPanel.test.ts
Root cause: vi.mock() factory functions referenced variables declared
outside the factory via vi.hoisted(), but vitest hoists vi.mock()
calls to the top of the file before variable initialization.
Solution: Define all mock implementations inline within the
vi.mock() factory functions, removing external variable references.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The vi.mock() callback was capturing mockManagerInstance at module
load time when it was undefined. Fixed by:
1. Changing mock to vi.fn() without return value
2. Adding mockReturnValue(mockManagerInstance) in beforeEach
This ensures the mock returns the properly configured instance when
getSemanticNarrativeManager() is called during panel construction.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Fixed 3 failing tests by correcting the mock setup:
- Moved mock function initialization from module-level to beforeEach
- This ensures fresh mock functions for each test run
- Added stub implementations for all interface methods
- Fixed mock manager instance creation per test
Tests now passing:
- "should refresh narrative from manager"
- "should generate narrative for worker and set it"
- "should generate aggregated narrative and set it"
The issue was that mock functions were defined globally at module
load time, causing them to be stale between test runs. Tests were
setting return values on stale function references, so mock calls
weren't being tracked properly.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Fixed the mock setup for the SemanticNarrativeManager to properly
allow mockReturnValue() to work on the manager methods.
The issue was that the mock class methods were defined as class
properties with vi.fn(() => null), which created new instances for
each test run. By defining the mock functions as module-level
variables and assigning them to the class methods, we ensure that
the same vi.fn() instances are used throughout, allowing tests to
properly configure return values with mockReturnValue().
This fixes 3 failing tests:
- "should refresh narrative from manager"
- "should generate narrative for worker and set it"
- "should generate aggregated narrative and set it"
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Re-verified all 59 tests in SemanticNarrativePanel.test.ts pass.
The issue was already fixed in commit 6237010 (bf-48nk).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
All 57 tests pass - failures were already fixed in commits 4839d48
and 797c9f1 (render() public, hide/show methods, test mock setup).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Re-verified all 57 tests passing on 2026-05-22 16:32.
No code changes required - all fixes were in previous commits.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Re-verified all 57 tests in FileContextPanel.test.ts pass.
The bead was already fixed by commits 4839d48 and 797c9f1.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Updated verification notes confirming all 57 tests pass. The test failures
described in the original bead were fixed by commits 797c9f1 and 4839d48,
which addressed mock setup issues and made render() public.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
- Make render() public so tests can call it directly without casting
- Change hide() to call this.render() instead of this.box.screen.render() directly
- This makes hide() consistent with show() and improves test compatibility
These changes address test failures in FileContextPanel.test.ts where:
- Tests call (panel as any).render() directly
- Tests expect mainBoxInstance.screen.render to be called after show/hide
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The original mock returned the same instance for all blessed.box() calls,
but the implementation creates 4 separate boxes (main, fileInfo, fileContent,
fileHistory). This caused tests to fail because they couldn't distinguish
which methods were called on which elements.
Changes:
- Make the mock return a new instance for each blessed.box() call
- Track all mock instances in a mockBoxes array
- Keep mainBoxInstance reference for tests that need the primary box
- Fix test expecting setContent on main box to use mockBoxes[1] (fileInfo)
All 57 FileContextPanel tests now pass.
Fixed test setup to correctly capture the mock instances that were
actually used during CommandPalette construction. The previous
approach of re-calling blessedMock.box() after clearing mocks
resulted in checking different mock instances than those used by
the actual palette object.
Changed from:
- Calling blessedMock.box() after construction (different instances)
To:
- Accessing blessedMock.box.mock.results[0]?.value (same instances)
This ensures tests correctly verify the behavior of the actual
CommandPalette instance, not mock instances that were never used
by the palette.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Confirmed that all issues from genesis bead description were already
fixed in previous runs:
- src/memoryProfiler.ts exists and fully implemented (255 lines)
- FileHeatmap treemap + timelapse: 31 tests pass
- SpanDag zoom/pan: tests pass
- All 2484 tests pass (4 skipped)
The original test failures were due to better-sqlite3 native module
version mismatch, not missing implementations.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
All tests were failing due to better-sqlite3 native module version
mismatch (NODE_MODULE_VERSION 137 vs 127). Rebuilt the module
with 'npm rebuild better-sqlite3' - all 2484 tests now pass.
No source code changes were needed; src/memoryProfiler.ts was already
present and complete.
The bead description claimed missing modules and features, but investigation
showed all implementation gaps were already closed. Test failures were due
to better-sqlite3 native module version mismatch (NODE_MODULE_VERSION 115
vs 137), resolved by rebuilding the module.
All 2484 tests now pass (4 skipped). No source code changes were needed.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
- Fix beadWorkspaceScanner.test.ts: make beforeEach async to use top-level await
- Fix parser.real-logs.integration.test.ts: change sequence test to non-decreasing (real logs have duplicate sequences)
- Fix FleetSummaryBar.test.tsx: correct expected separator count from 5 to 4
All 2484 tests now pass (was 89 failed due to wrong Node.js version + 3 test bugs).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Phase 9 UI change 2. Updated WorkerGrid component to display bead progress
instead of raw event count. Shows current bead when WORKING state.
- Added beadsCompleted and currentBead fields to frontend WorkerInfo type
- Worker card now shows "bead: bf-5r22 / 31 completed" when WORKING
- Shows "31 completed" when not WORKING or no current bead
- Removed eventCount display from worker card UI
- Updated tests to cover new display format
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Phase 9 UI change 1: Add always-visible summary line at top of web dashboard.
- New FleetSummaryBar component showing:
- N WORKING (workers in WORKING needleState)
- N SELECTING (workers in SELECTING needleState)
- N EXHAUSTED (workers in EXHAUSTED_IDLE or STOPPED)
- N beads today (sum of beadsCompleted from workers)
- N stuck (count of workers with stuck=true)
- Wired into App.tsx above WorkerGrid
- Comprehensive frontend tests covering all states and reactivity
- CSS styling with color-coded states
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Phase 9 implementation: Bead workspace scanner and project breakdown.
- Add beadWorkspaceScanner.ts to scan .beads/issues.jsonl files
- Count CLOSED beads per project, deriving project from bead id prefix
- Use close_reason/closed_at/assignee for productivity tracking
- Add configurable workspace list in config.ts (WorkspaceConfig interface)
- Extend GET /api/productivity to add byProject array
- Add By Project section to ProductivityPanel React component
- Add tests for bead workspace scanner
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The working-tree issues.jsonl had been wiped (db reset); this restores the
authoritative 153-bead history and adds the Phase 9 Productivity Analytics
gap beads created 2026-05-22:
bf-2wf epic: Phase 9 remaining gaps
bf-60j currentBead field on WorkerInfo (completed by worker)
bf-3xp bead workspace scanner + /api/productivity project breakdown
bf-4f3 fleet summary bar component
bf-3t8 worker-card enrichment (blocked by bf-60j)
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Add currentBead field to WorkerInfo in src/types.ts and maintain it in
src/store.ts: set it from bead.claim.succeeded (event.bead) and clear it
(null) on bead.released. Foundation for the worker-card current-bead
display. Update src/store.test.ts to cover claim->set and release->clear.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Adds GET /api/productivity returning daily bead completion counts (last 30
days) from bead.released/release_success events and a worker leaderboard
sorted by beadsCompleted. Adds a Productivity tab in the web UI with a 14-day
SVG bar chart and a worker leaderboard table.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Modern NEEDLE emits bead.released with reason=release_success instead of
legacy bead.completed. Update store.ts and workerAnalytics.ts to handle
this event: increment beadsCompleted, set status to idle, and clear
activeBead/activeFiles/activeDirectories.
Also fixes workerAnalytics trackBeadEvent to use explicit isReleasedSuccess
check instead of broad msg.includes('success') which could false-positive.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
WorkerGrid now sorts by WORKING > CLAIMING > SELECTING > BOOTING >
CLOSING > EXHAUSTED_IDLE > STOPPED and filters out test workers
(test-*, claude-test-*, nonexistent-*, needle-test, strand-runner,
-test-worker) by default. App header gets a toggle button to show/hide
test workers. EXHAUSTED_IDLE added to NeedleState type and propagated
to WorkerDetail and WorkerGrid state maps.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
