The createClosedIssue helper was setting ClosedAt to time.Now(), but the
eligibility check uses '<= datetime(now, -0 days)'. This created a race
condition where if nanoseconds elapsed between issue creation and the
query, the issue would fail eligibility checks.
Changed to now.Add(-1*time.Second) to ensure issues are always eligible
when compact_tier1_days is set to 0 for testing.
Fixes bd-18
Amp-Thread-ID: https://ampcode.com/threads/T-bc303dd9-f0e7-4d52-8908-e433d429ac4a
Co-authored-by: Amp <amp@ampcode.com>
Tests were connecting to test daemon but daemon routed to production DB via
findDatabaseForCwd(). Fixed by ensuring tests use isolated .beads directories
and change working directory to tmpDir.
Changes:
- bench_test.go: Added .beads subdir, chdir, and client.dbPath to setupBenchServer
- bench_test.go: Set dbPath for goroutine clients in BenchmarkConcurrentAgents
- comments_test.go: Refactored to use setupTestServer
- version_test.go: Fixed 4 tests to use setupTestServerIsolated with proper isolation
- rpc_test.go: Added setupTestServerIsolated() helper for custom test setup
Verified: RPC test suite runs with no database pollution (151→151 issues)
Amp-Thread-ID: https://ampcode.com/threads/T-348b7ba8-4292-4ed3-b143-0ad07d226c21
Co-authored-by: Amp <amp@ampcode.com>
Add //go:build bench tag to bench_test.go so benchmarks NEVER run
during normal 'go test'. They've polluted production DB multiple times.
Benchmarks only run with: go test -tags=bench -bench=.
Cleaned up 1013 test pollution issues (Agent X Issue Y, Version tests).
Back to 18 real open issues.
Amp-Thread-ID: https://ampcode.com/threads/T-5ad38d33-28ba-4f47-997a-b0d7e0331c26
Co-authored-by: Amp <amp@ampcode.com>
- Added OpExport/OpImport to RPC protocol
- Implemented handleExport() in daemon to export via RPC
- Modified sync command to use daemon export when available
- Prevents nil pointer dereference when daemon is running
- Falls back to direct mode if daemon unavailable
bd-50: Verified all counter sync fixes already implemented
- Import calls SyncAllCounters() after batch operations
- Delete operations sync counters properly
- Renumber resets and syncs counters correctly
- Daemon cache detects external DB changes via mtime
- Added comprehensive tests: TestCounterSyncAfterImport
bd-51: Cleaned up test pollution from production database
- Deleted bd-52 through bd-58 (manual test issues)
- Root cause was user error, not auto-flush bug
- Auto-flush working as designed
- Go tests properly isolated in temp directories
Amp-Thread-ID: https://ampcode.com/threads/T-9dcbc4bb-76fb-4696-a3f4-4af560da6d6c
Co-authored-by: Amp <amp@ampcode.com>
- Call SyncAllCounters() after DeleteIssue and DeleteIssues
- Change SyncAllCounters to use excluded.last_id (allows counter to decrease)
- Delete orphaned counter rows when no issues remain for a prefix
- Add comprehensive tests in counter_sync_test.go
Fixes the issue where deleting issues left counters at high values, causing
new issues to skip IDs. Now counters accurately reflect the max existing ID.
Closes bd-49
Amp-Thread-ID: https://ampcode.com/threads/T-c3bdb8b9-d67b-4de5-901e-7ea76fc9e399
Co-authored-by: Amp <amp@ampcode.com>
- Created TestDatabaseIsolation in internal/rpc/isolation_test.go
- Single-issue test verifies test DB isolation works correctly
- Prevents repeat of stress test pollution incident
- Test passes cleanly without polluting production database
Closes bd-44
Amp-Thread-ID: https://ampcode.com/threads/T-b3e81b42-2498-45b5-9c8b-beb27f7e9550
Co-authored-by: Amp <amp@ampcode.com>
- Deleted internal/rpc/stress_test.go (was polluting production DB)
- Removed 1002 test issues (Agent X Issue Y, Multi-ID tests)
- Back to 99 real issues from 1101 total
- Closes bd-48
The TestStressNoUniqueConstraintViolations test was creating 1000 test
issues in the project's real database instead of its temp database.
Root cause: The RPC client automatically sets req.Cwd to os.Getwd() when
not explicitly provided. The server's getStorageForRequest() then uses
findDatabaseForCwd() to locate and open the database for that working
directory. Since the test ran from the project directory, it discovered
and used .beads/bd.db instead of the temp database.
Fix: Change working directory to the temp directory before running the
test, so the client's automatic Cwd detection points to the temp .beads/
directory instead of the project's.
Also ensures .beads subdirectory exists in temp dir for proper discovery.
The daemon's handleCreate function was ignoring createArgs.Labels and
createArgs.Dependencies, causing labels and deps specified during issue
creation to be lost when using daemon mode.
Added label and dependency processing after issue creation to match
the direct mode behavior in cmd/bd/main.go.
Fixes: #101
Amp-Thread-ID: https://ampcode.com/threads/T-cc80f8d7-264c-446e-b34f-79c4a7713b77
Co-authored-by: Amp <amp@ampcode.com>
- Add dbMtime field to StorageCacheEntry to track DB file modification time
- Check mtime on cache hits and evict stale entries if DB changed externally
- Close and reopen storage when external modifications detected
- Fixes issue where daemon served stale data after direct DB operations
Amp-Thread-ID: https://ampcode.com/threads/T-631d5cca-0b26-47cb-b633-118b788483cf
Co-authored-by: Amp <amp@ampcode.com>
- Add ExpectedDB field to RPC Request
- Server validates client's expected DB matches daemon's DB
- Return clear error on mismatch with both paths
- Old clients (no ExpectedDB) still work with warning
- Add Path() method to storage.Storage interface
- Tests verify cross-database connections rejected
Prevents database pollution when client connects to wrong daemon.
Amp-Thread-ID: https://ampcode.com/threads/T-c4454192-39c6-4c67-96a9-675cbfc4db92
Co-authored-by: Amp <amp@ampcode.com>
- Remove ~/.beads/bd.sock fallback in getSocketPath()
- Always return local socket path (.beads/bd.sock)
- Add migration warning if old global socket exists
- Update AGENTS.md to remove global daemon references
- Document breaking change in CHANGELOG.md
- Fix test isolation: tests now use temp .beads dir and chdir
Prevents cross-project daemon connections and database pollution.
Each project must use its own local daemon.
Amp-Thread-ID: https://ampcode.com/threads/T-c4454192-39c6-4c67-96a9-675cbfc4db92
Co-authored-by: Amp <amp@ampcode.com>
- Add --id flag accepting comma-separated IDs
- Implements ID filtering at CLI, RPC, and storage layers
- Normalizes IDs (trim, dedupe, remove empty) like labels
- Guards against excessive ID lists (max 1000)
- Works with other filters (status, priority, etc.)
Closes bd-200
Amp-Thread-ID: https://ampcode.com/threads/T-377464f2-1e7f-46f9-b23e-1e3cfd611061
Co-authored-by: Amp <amp@ampcode.com>
Use simpler approach for bd-190 merge feature:
- Close merged issues with reason 'Merged into bd-X'
- No schema changes or migrations needed
- Parseable close reason is cleaner than separate field
Also updated merge epic child issues with simplified design.
- Use sql.NullString to scan nullable assignee column
- Prevents 'converting NULL to string is unsupported' error
- Only set assignee if value is valid (not NULL)
- Added OpCompact and OpCompactStats operation constants
- Added CompactArgs, CompactStatsArgs, and response types to RPC protocol
- Implemented handleCompact and handleCompactStats in RPC server
- Updated compact command to use RPC when daemon is available
- Fixed RPC client to include Cwd for proper database routing
- Compact now works in daemon mode with --no-daemon flag
Amp-Thread-ID: https://ampcode.com/threads/T-87885d07-80ad-466d-9ffb-cc96fab4853f
Co-authored-by: Amp <amp@ampcode.com>
- Add dbPath field to SQLiteStorage to track database file path
- Create derivePrefixFromPath() helper to extract prefix from filename
- Update ID generation in CreateIssue() and generateBatchIDs() to use filename fallback
- Fix tests to explicitly set issue_prefix config for bd- prefixed tests
When config doesn't have issue_prefix set, bd now correctly derives it from
the database filename (e.g., wy-.db -> wy) instead of always defaulting to 'bd'.
Fixes: bd-179
- Add --label flag for AND filtering (must have ALL labels)
- Add --label-any flag for OR filtering (must have AT LEAST ONE label)
- Add normalizeLabels() helper to trim, dedupe, and clean inputs
- Fix RPC title filtering parity bug (forward via Query field)
- Add comprehensive tests for label filtering including combined AND+OR
- Update documentation in README and CHANGELOG
- Improve flag help text to clarify combined semantics
Closes bd-161
- Populate labels in bd list human-readable output (direct mode)
- Populate labels in bd list JSON output (direct and daemon modes)
- Populate labels in daemon RPC handlers (handleList, handleShow)
- handleShow now returns full IssueDetails with labels/deps/dependents
- Labels displayed with 'Labels: [label1, label2]' format
Closes bd-164
Amp-Thread-ID: https://ampcode.com/threads/T-30cd607d-c509-4a8c-9cac-c2aea2ad75c6
Co-authored-by: Amp <amp@ampcode.com>
- Tests for daemon/client version checking
- Covers all compatibility scenarios (major/minor/patch versions)
- Tests legacy client support (empty version)
- Tests health check version reporting
- Tests ping/health bypass version checks
- All 13 test cases pass
Closes bd-160
Amp-Thread-ID: https://ampcode.com/threads/T-30cd607d-c509-4a8c-9cac-c2aea2ad75c6
Co-authored-by: Amp <amp@ampcode.com>
- Updated label CLI commands to support both daemon and direct modes
- Added label fetching to GetIssue() and scanIssues() methods
- All label operations (add, remove, list, list-all) work with daemon
- Closed bd-162 (label CLI commands), bd-166 (duplicate), bd-141 (daemon support)
Amp-Thread-ID: https://ampcode.com/threads/T-4858f62e-ad06-4cc7-ad05-17ee76861f86
Co-authored-by: Amp <amp@ampcode.com>
Add WaitReady() channel to RPC server that signals when the socket is
listening and ready to accept connections. Previously daemon startup
waited a fixed 2 seconds which could fail if the server took longer.
Changes:
- Add readyChan to Server struct
- Signal ready after listener bind completes
- Update daemon startup to wait on WaitReady() channel
- Increase timeout to 5s with proper signaling
This fixes multi-repo daemon routing test failures where daemon would
start but not be ready to handle requests within the timeout window.
Amp-Thread-ID: https://ampcode.com/threads/T-675a2db5-b1b3-480d-a108-b003d8139d08
Co-authored-by: Amp <amp@ampcode.com>
Fixes timestamp scanning error reported in GH-88 where DATETIME columns
were being returned as strings instead of time.Time on macOS 13.5.
Root cause: modernc.org/sqlite driver doesn't recognize mattn-style DSN
parameters (_journal_mode, _foreign_keys). When these incompatible
parameters are present, the driver ignores _time_format=sqlite on some
platforms, causing DATETIME values to remain as strings.
Solution: Use modernc's native _pragma syntax for all database options:
- Changed _journal_mode=WAL to _pragma=journal_mode(WAL)
- Changed _foreign_keys=ON to _pragma=foreign_keys(ON)
- Kept _pragma=busy_timeout(30000) and _time_format=sqlite
This ensures all parameters are properly recognized and DATETIME columns
are automatically parsed to time.Time across all platforms.
Fixes#88
Amp-Thread-ID: https://ampcode.com/threads/T-44d1817a-3709-4f1d-a27a-78bb2fa4d3dc
Co-authored-by: Amp <amp@ampcode.com>
Fixes bd-160
The race was between Start() writing s.listener and Stop() reading it.
Now all listener access is protected by the server mutex:
- Start() stores listener under lock after creation
- Accept loop reads listener under RLock
- Stop() closes listener under lock
All RPC tests now pass with -race flag.
- Add --max-depth/-d flag with default of 50
- Wire flag through to store.GetDependencyTree()
- Add input validation (must be >= 1)
- Show inline '… [truncated]' markers on truncated nodes
- Update truncation warning to show actual depth used
- Add comprehensive tests (truncation, default depth, boundary cases)
- Update CLI docs and reference
Thanks to @yashwanth-reddy909 for the initial implementation in PR #87.
This commit completes the feature with full wiring, validation, tests, and docs.
Amp-Thread-ID: https://ampcode.com/threads/T-c439b09c-cff2-48d9-8988-cf9353f0d32e
Co-authored-by: Amp <amp@ampcode.com>
Implemented bd-150: Improve daemon fallback visibility and user feedback
- Added DaemonStatus struct to track connection state
- Enhanced BD_DEBUG logging with detailed diagnostics and timing
- Added BD_VERBOSE mode with actionable warnings when falling back
- Implemented health checks before using daemon
- Clear fallback reasons: connect_failed, health_failed, auto_start_disabled, auto_start_failed, flag_no_daemon
- Updated documentation
Implemented bd-151: Add version compatibility checks for daemon RPC protocol
- Added ClientVersion field to RPC Request struct
- Client sends version (0.9.10) in all requests
- Server validates version compatibility using semver:
- Major version must match
- Daemon >= client for backward compatibility
- Clear error messages with directional hints (upgrade daemon vs upgrade client)
- Added ClientVersion and Compatible fields to HealthResponse
- Implemented 'bd version --daemon' command to check compatibility
- Fixed batch operations to propagate ClientVersion for proper checks
- Updated documentation with version compatibility section
Code review improvements:
- Propagate ClientVersion in batch sub-requests
- Directional error messages based on which side is older
- Made ServerVersion a var for future unification
Amp-Thread-ID: https://ampcode.com/threads/T-b5fe36b8-c065-44a9-a55b-582573671609
Co-authored-by: Amp <amp@ampcode.com>
- Implements bd stale command to show issues with execution_state where executor is dead/stopped
- Adds --release flag to automatically release orphaned issues
- Adds --threshold flag to customize heartbeat staleness threshold (default: 300s/5min)
- Handles missing executor instances (LEFT JOIN) for cases where executor was deleted
- Adds QueryContext and BeginTx helper methods to SQLiteStorage for advanced queries
- Fixes ExternalRef comparison bug in import_shared.go (pointer vs string)
- Removes unused imports in import.go
Resolves vc-124
Improvements based on oracle code review:
- Move socket cleanup AFTER lock acquisition (prevents unlinking live sockets)
- Add PID liveness check before removing stale socket
- Add stale lock detection with retry mechanism
- Tighten directory permissions to 0700 for security
- Improve socket readiness probing with shorter timeouts
- Make removeOldSocket() ignore ENOENT errors
Fixes race condition where socket could be removed during daemon startup window,
potentially orphaning a running daemon process.
Amp-Thread-ID: https://ampcode.com/threads/T-63542c60-b5b9-4a34-9f22-415d9d7e8223
Co-authored-by: Amp <amp@ampcode.com>
- Add OpHealth RPC operation to protocol
- Implement handleHealth() with DB ping and 1s timeout
- Returns status (healthy/degraded/unhealthy), uptime, cache metrics
- Update TryConnect() to use health check instead of ping
- Add 'bd daemon --health' CLI command with JSON output
- Track cache hits/misses for metrics
- Unhealthy daemon triggers automatic fallback to direct mode
- Health check completes in <2 seconds
Amp-Thread-ID: https://ampcode.com/threads/T-1a4889f3-77cf-433a-a704-e1c383929f48
Co-authored-by: Amp <amp@ampcode.com>
Implements bd-9: Allow users to view all paths through diamond dependencies
without deduplication. Useful for debugging complex dependency structures.
Changes:
- Added --show-all-paths flag to bd dep tree command
- Updated GetDependencyTree interface to accept showAllPaths parameter
- Modified deduplication logic to be conditional on flag
- Updated tests to pass new parameter
Amp-Thread-ID: https://ampcode.com/threads/T-43807dd5-8732-49ad-a839-cdb5dae70c35
Co-authored-by: Amp <amp@ampcode.com>
- bd-170: Implement hybrid sorting for ready work (recent 48h first, then oldest)
- bd-87: Use safer null-byte placeholders in ID remapping
- bd-92: Make auto-flush debounce configurable via BEADS_FLUSH_DEBOUNCE
- bd-171: Fix nil pointer dereference in renumber command
- Delete spurious test issues (bd-7, bd-130-134)
- Renumber database from 171 down to 144 issues
- bd-159: Global daemon now runs in routing mode without opening DB
- bd-158: Set socket permissions to 0600 for security
- bd-160: Reject --auto-commit/--auto-push with --global
- bd-157: Verified stale socket cleanup (already working)
- bd-56: Closed as won't-do (cycle prevention is better)
- bd-73: Multi-repo support complete
- Add DeleteIssues() method in sqlite.go for atomic batch deletion
- Support multiple issue IDs as arguments or from file
- Add --from-file flag to read IDs from file (supports comments)
- Add --dry-run mode for safe preview without deleting
- Add --cascade flag for recursive deletion of dependents
- Add --force flag to orphan dependents instead of failing
- Pre-collect connected issues before deletion for text reference updates
- Add orphan deduplication to prevent duplicate IDs
- Add rows.Err() checks in all row iteration loops
- Full transaction safety - all deletions succeed or none do
- Comprehensive statistics tracking (deleted, dependencies, labels, events)
- Update README and CHANGELOG with batch deletion docs
Fixed critical code review issues:
- Dry-run mode now properly uses dryRun parameter instead of deleting data
- Text references are pre-collected before deletion so they update correctly
- Added orphan deduplication and error checks
- Updated defer rollback pattern per Go best practices
- Added per-request storage routing in daemon server
- Server now supports Cwd field in requests for database discovery
- Tree-walking to find .beads/*.db from any working directory
- Storage caching for performance across requests
- Created Python daemon client (bd_daemon_client.py)
- RPC over Unix socket communication
- Implements full BdClientBase interface
- Auto-discovery of daemon socket from working directory
- Refactored bd_client.py with abstract interface
- BdClientBase abstract class for common interface
- BdCliClient for CLI-based operations (renamed from BdClient)
- create_bd_client() factory with daemon/CLI fallback
- Backwards-compatible BdClient alias
Next: Update MCP server to use daemon client when available
- Add 'bd epic status' to show epic completion with child progress
- Add 'bd epic close-eligible' to bulk-close completed epics
- Add GetEpicsEligibleForClosure() storage method
- Update 'bd stats' to show count of epics ready to close
- Add EpicStatus type for tracking epic/child relationships
- Support --eligible-only, --dry-run, and --json flags
- Fix golangci-lint config version requirement
Addresses GitHub issue #62 - epics now have visibility and
management tools for closure when all children are complete.
Amp-Thread-ID: https://ampcode.com/threads/T-e8ac3f48-f0cf-4858-8e8f-aace2481c30d
Co-authored-by: Amp <amp@ampcode.com>
Use _time_format=sqlite parameter in modernc.org/sqlite connection
string to ensure DATETIME columns use SQLite's native time format
(format 7 with timezone) instead of Go's default String() format.
This improves compatibility with SQLite's date/time functions and
ensures consistent time representation across the database.
The counter wasn't being properly reset after renumbering because
SyncAllCounters uses MAX(old, new) which kept higher values from
deleted issues.
Solution: Add ResetCounter() method to delete the counter entry,
then SyncAllCounters recreates it from the actual max ID in database.
Now after renumbering 108 issues to bd-1..bd-108, the counter is
correctly set to 108 and next issue will be bd-109.
