The --protect-left-snapshot mechanism was protecting ALL local issues
by ID alone, ignoring timestamps. This caused newer remote changes to
be incorrectly skipped during cross-worktree sync.
Changes:
- Add BuildIDToTimestampMap() to SnapshotManager for timestamp-aware
snapshot reading
- Change ProtectLocalExportIDs from map[string]bool to map[string]time.Time
- Add shouldProtectFromUpdate() helper that compares timestamps
- Only protect if local snapshot is newer than incoming; allow update
if incoming is newer
This fixes data loss scenarios where:
1. Main worktree closes issue at 11:31
2. Test worktree syncs and incorrectly skips the update
3. Test worktree then pushes stale open state, overwriting mains changes
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Strip (bd-xxx), (gt-xxx) suffixes from code comments and changelog
entries. The descriptions remain meaningful without the ephemeral
issue IDs.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The buildAllowedPrefixSet function now parses the allowed_prefixes config
and includes those prefixes when validating imports. This allows mol-*
prefixes to be imported without errors when configured.
Config example: bd config set allowed_prefixes "gt-,mol-"
Comment timestamps were being overwritten with CURRENT_TIMESTAMP during
import, causing infinite sync loops between hosts as each import would
update timestamps.
Added ImportIssueComment() method that accepts and preserves the original
timestamp from JSONL, and updated importComments() to use it.
Closes#735
The prefix mismatch detection was using ExtractIssuePrefix() which
tries to guess the prefix by analyzing suffix patterns. This failed
for multi-hyphen prefixes like "asianops-audit-" when issue IDs had
word-like suffixes (e.g., "asianops-audit-test") - the heuristic
would fall back to the first hyphen and report "asianops-" as the
prefix.
Fixed by checking directly if the issue ID starts with the configured
prefix using strings.HasPrefix(). This is more reliable than guessing
since we know the expected prefix from the database config.
Added test case TestImportMultiHyphenPrefix to prevent regression.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Fixed two code paths where pinned=false from JSONL would overwrite
existing pinned=true in database:
- importer.go: Only update pinned if explicitly true in JSONL
- multirepo.go: Use COALESCE to preserve existing pinned value
Added tests for pinned field preservation.
Note: Bug may have additional code path - investigation ongoing.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The 'bd pin' command was failing with "invalid field for update: pinned"
because the pinned field was missing from allowedUpdateFields.
Fixes:
- Add 'pinned' to allowedUpdateFields in queries.go
- Update importer to include pinned field in updates during import
- Add equalBool comparator for IssueDataChanged to detect pinned changes
- Fix stats query to count pinned=1 instead of status='pinned'
- Fix pinIndicator in list.go to check issue.Pinned instead of status
This unblocks gt mail --pinned functionality.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* Add Linear integration CLI with sync and status commands
- Add `bd linear sync` for bidirectional issue sync with Linear
- Add `bd linear status` to show configuration and sync state
- Stub pull/push functions pending GraphQL client (bd-b6b.2)
* Implement Linear GraphQL client with full sync support
- Add LinearClient with auth, fetch, create, update methods
- Implement pull/push operations with Beads type mapping
- Clean up redundant comments and remove unused code
* Add configurable data mapping and dependency sync for Linear
- Add LinearMappingConfig with configurable priority/state/label/relation maps
- Import parent-child and issue relations as Beads dependencies
- Support custom workflow states via linear.state_map.* config
* Add incremental sync support for Linear integration
- Add FetchIssuesSince() method using updatedAt filter in GraphQL
- Check linear.last_sync config to enable incremental pulls
- Track sync mode (incremental vs full) in LinearPullStats
* feat(linear): implement push updates for existing Linear issues
Add FetchIssueByIdentifier method to retrieve single issues by identifier
(e.g., "TEAM-123") for timestamp comparison during push.
Update doPushToLinear to:
- Fetch Linear issue to get internal ID and UpdatedAt timestamp
- Compare timestamps: only update if local is newer
- Build update payload with title, description, priority, and state
- Call UpdateIssue for issues where local has newer changes
Closes bd-b6b.5
* Implement Linear conflict resolution strategies
- Add true conflict detection by fetching Linear timestamps via API
- Implement --prefer-linear resolution (re-import from Linear)
- Implement timestamp-based resolution (newer wins as default)
- Fix linter issues: handle resp.Body.Close() and remove unused error return
* Add Linear integration tests and documentation
- Add comprehensive unit tests for Linear mapping (priority, state, labels, relations)
- Update docs/CONFIG.md with Linear configuration reference
- Add examples/linear-workflow guide for bidirectional sync
- Remove AI section header comments from tests
* Fix Linear GraphQL filter construction and improve test coverage
- Refactor filter handling to combine team ID into main filter object
- Add test for duplicate issue relation mapping
- Add HTTP round-trip helper for testing request payload validation
* Refactor Linear queries to use shared constant and add UUID validation
- Extract linearIssuesQuery to deduplicate FetchIssues/FetchIssuesSince
- Add linearMaxPageSize constant and UUID validation with regex
- Expand test coverage for new functionality
* Refactor Linear integration into internal/linear package
- Extract types, client, and mapping logic from cmd/bd/linear.go
- Create internal/linear/ package for better code organization
- Update tests to work with new package structure
* Add linear teams command to list available teams
- Add FetchTeams GraphQL query to Linear client
- Refactor config reading to support daemon mode
- Add tests for teams listing functionality
* Refactor Linear config to use getLinearConfig helper
- Consolidate config/env var lookup using getLinearConfig function
- Add LINEAR_TEAM_ID environment variable support
- Update error messages to include env var configuration options
* Add hash ID generation and improve Linear conflict detection
- Add configurable hash ID mode for Linear imports (matches bd/Jira)
- Improve conflict detection with content hash comparison
- Enhance conflict resolution with skip/force update tracking
* Fix test for updated doPushToLinear signature
- Add missing skipUpdateIDs parameter to test call
The early ID deduplication in the main loop (seenIDs check) already
prevents duplicate IDs from reaching newIssues. The pre-batch
deduplication was dead code that could never fire.
Keeping two layers:
1. Early dedup in main loop - catches duplicates in JSONL input
2. INSERT OR IGNORE at DB level - safety net for all code paths
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- IsExpired(): Negative TTL means immediately expired (for --hard mode)
- IsExpired(): ClockSkewGrace only added for TTLs > 1 hour
- bd cleanup --hard: Use negative TTL to prune freshly created tombstones
- bd delete --hard: New flag to immediately prune tombstones from JSONL
- Import: Add early tombstone check before all phases to prevent resurrection
The early tombstone check prevents ghost issues from being created when
tombstones exist in the DB. However, a deeper git merge issue (bd-ncwo)
can still cause resurrection when remote's status:closed wins the merge.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- bd-6pni: Auto-filter tombstoned issues with mismatched prefixes during
import instead of failing. Tombstones from contributor PRs with different
test prefixes are pollution and safe to ignore.
- bd-ffr9: Stop recreating deletions.jsonl after tombstone migration.
Added IsTombstoneMigrationComplete() check to all code paths that write
to the legacy deletions manifest.
- bd-admx: Fix perpetual "JSONL file hash mismatch" warning. Now clears
both export_hashes AND jsonl_file_hash when mismatch detected, so the
warning doesn't repeat.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- bd-8v5o: When doctor --fix hydrates issues from git history, also
remove them from the deletions manifest to prevent perpetual skip
warnings during sync
- bd-wsqt: Remove verbose per-issue "Skipping bd-xxx" messages during
sync. Caller already shows summary of skipped issues.
Added RemoveDeletions() function to deletions package with tests.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
When importing issues, if an incoming issue has the same content hash as an
existing issue but a DIFFERENT prefix, this should not be treated as a rename.
Cross-prefix content matches occur when importing issues from other projects
that happen to have identical content.
Previously, the importer would call handleRename which tries to create an issue
with the incoming prefix, failing prefix validation ("does not match configured
prefix" error).
The fix checks if prefixes differ before calling handleRename:
- Same prefix, different ID suffix → true rename, call handleRename
- Different prefix → skip incoming issue, keep existing unchanged
Added test: TestImportCrossPrefixContentMatch reproduces the bug scenario
where alpha-* issues exist but beta-* issues are imported with same content.
* bd-ckej: fix orphan skip count mismatch on fresh import
When OrphanSkip mode is used during import and a child issue's parent doesn't
exist, the issue ID was cleared to '' but then regenerated anyway in
GenerateBatchIssueIDs, causing it to be created in the database. This resulted
in a count mismatch: JSONL had 824 issues but only 823 were in the database (one
orphan was counted but not created).
Fix: Filter out orphaned issues with empty IDs before batch creation and track
them in result.Skipped so the count stays accurate.
* test: add TestImportOrphanSkip_CountMismatch for bd-ckej
Adds comprehensive test that verifies orphaned issues are properly skipped
during import when orphan_handling=OrphanSkip and parent doesn't exist.
Also improves the fix to pre-filter orphaned issues before batch creation,
ensuring they're not inserted then have IDs cleared (preventing count
mismatches).
---------
Co-authored-by: Amp <amp@example.com>
GH#464: Add safety guards to prevent deletion of open/in_progress issues during sync:
- Safety guard in git-history-backfill (importer.go)
- Safety guard in deletions manifest processing
- Warning when uncommitted changes detected before pull (daemon_sync.go)
- Enhanced repo ID mismatch error message
GH#545: Fix bd blocked to show status=blocked issues (sqlite/ready.go):
- Changed from INNER JOIN to LEFT JOIN to include issues without dependencies
- Added WHERE clause to include both status=blocked AND dependency-blocked issues
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The importer was not seeing tombstones when building the dbByID map,
causing it to treat tombstone IDs as "new" issues. This led to UNIQUE
constraint violations during INSERT.
- Include tombstones in SearchIssues call (IncludeTombstones: true)
- Skip tombstones when matching by ID instead of trying to update them
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Fix sync bug where newly created issues were incorrectly tombstoned during bd sync.
The root cause was git-history-backfill finding issues in local commits on the sync branch, then tombstoning them when they weren't in the merged JSONL. The fix protects issues from the left snapshot (local export) from git-history-backfill.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Changed Priority from hardcoded 2 to 0 (unset) to distinguish legacy tombstones
from user-set values. IssueType remains TypeTask as empty fails validation.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Track legacy deletions.jsonl entries converted to tombstones during import:
- Add Result.ConvertedToTombstone counter
- Add Result.ConvertedTombstoneIDs for the converted IDs
- Update test to verify the new counter
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Update import/export to handle tombstones for deletion sync propagation:
Exporter:
- Include tombstones in JSONL output by setting IncludeTombstones: true
- Both single-repo and multi-repo exports now include tombstones
Importer:
- Tombstones from JSONL are imported as-is (they're issues with status=tombstone)
- Legacy deletions.jsonl entries are converted to tombstones via convertDeletionToTombstone()
- Non-tombstone issues in deletions manifest are still skipped (backward compat)
- purgeDeletedIssues() now creates tombstones instead of hard-deleting
This is Phase 2 of the tombstone implementation (bd-dli design), enabling
inline soft-delete tracking for cross-clone deletion synchronization.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
When a clone gets reset (git reset --hard origin/main), the
git-history-backfill logic was incorrectly marking ALL issues as
deleted since they appeared in git history but not in the current
JSONL. This caused the entire database to be purged.
Fix:
- Add 50% threshold check: abort if git-history-backfill would delete
more than 50% of issues (likely a reset scenario, not deletions)
- Add warning when >10 issues would be deleted via backfill
- Print helpful message about manual deletion if needed
Test:
- Added TestMassDeletionSafetyGuard that simulates JSONL reset and
verifies the safety guard prevents mass deletion
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
When importing JSONL that contains issues in the deletions manifest,
import now:
- Filters out deleted issues before import
- Prints per-issue warning with deletion details (date, actor)
- Shows count of skipped issues in summary
- Suggests --ignore-deletions flag to force import
The new --ignore-deletions flag allows importing issues that are in the
deletions manifest, useful for recovering accidentally deleted issues.
Fixes bd-4zy
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
When auto-importing issues from JSONL, issues with different prefixes
(e.g., gt-1 vs gastown-) would fail validation and cause an infinite
loop of failed migrations.
The fix adds SkipPrefixValidation option to CreateIssuesWithFullOptions
which propagates through EnsureIDs to skip prefix validation for issues
that already have IDs during import. This allows importing issues with
any prefix while still validating new issues created interactively.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Fixes bd-0b2: The git history backfill mechanism was causing data loss
during JSONL filename migrations (beads.jsonl → issues.jsonl). When issues
existed in the old filename's git history, the backfill incorrectly treated
them as "deleted" and purged them from the database.
Changes:
- Add NoGitHistory field to importer.Options and ImportOptions structs
- Modify purgeDeletedIssues() to skip git history check when flag is set
- Add --no-git-history flag to bd import command
- Add --no-git-history flag to bd sync command
- Update purge_test.go to pass Options argument
Usage:
bd import -i .beads/issues.jsonl --no-git-history
bd sync --no-git-history
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Change default JSONL filename from beads.jsonl to issues.jsonl
- Add bd doctor check and fix to auto-migrate legacy beads.jsonl configs
- Update FindJSONLPath to prefer issues.jsonl over beads.jsonl
- Add CheckLegacyJSONLConfig and CheckLegacyJSONLFilename checks
- Add LegacyJSONLConfig fix to rename files and update config
- Update .gitattributes to reference issues.jsonl
- Fix tests to expect new canonical filename
- Add bd-6xd to v0.25.1 release notes
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Completes the deletion propagation epic (bd-imj) with all 9 subtasks:
- Cross-clone deletion propagation via deletions.jsonl
- bd deleted command for audit trail
- Auto-compact during sync (opt-in)
- Git history fallback with timeout and regex escaping
- JSON output for pruning results
- Integration tests for deletion scenarios
- Documentation in AGENTS.md, README.md, and docs/DELETIONS.md
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Add Count function to deletions package for fast line counting
- Add maybeAutoCompactDeletions to sync (opt-in via deletions.auto_compact config)
- Fix regex escaping in batchCheckGitHistory (bd-bgs)
- Add 30s timeout to git history commands (bd-f0n)
- Use git rev-parse --show-toplevel for proper repo root detection (bd-bhd)
- Add tests for Count and auto-compact functionality
Closes: bd-qsm, bd-bgs, bd-f0n, bd-bhd
Implements two P1 tasks for the deletions manifest epic:
bd-v2x: Add deletions pruning to bd compact
- PruneDeletions function removes records older than retention period
- Default retention: 7 days (configurable via metadata.json)
- CLI --retention flag for override
- Atomic file rewrite prevents corruption
- Called automatically during all compact operations
bd-pnm: Add git history fallback for pruned deletions
- Catches deletions where manifest entry was pruned
- Uses git log -S to search for ID in JSONL history
- Batches multiple IDs for efficiency (git -G regex)
- Self-healing: backfills manifest on hit
- Conservative: keeps issue if git check fails (shallow clone)
Tests added for both features with edge cases covered.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
When importing JSONL after merges that include deletions, FK constraint
violations can occur if an issue references a deleted issue. Previously,
import would fail completely. Now it continues and reports skipped dependencies.
Changes:
- Add SkippedDependencies field to Result/ImportResult structs
- Update importDependencies() to detect FK violations using IsForeignKeyConstraintError()
- Log warnings for each skipped dependency with issue IDs and type
- Continue importing remaining dependencies instead of failing
- Display summary of all skipped dependencies at end of import
Example output:
Warning: Skipping dependency due to missing reference: bd-b → bd-a (blocks)
⚠️ Warning: Skipped 2 dependencies due to missing references:
- bd-b → bd-a (blocks)
- bd-c → bd-a (parent-child)
This can happen after merges that delete issues referenced by other issues.
The import continued successfully - you may want to review the skipped dependencies.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Complete implementation of signal-aware context propagation for graceful
cancellation across all commands and storage operations.
Key changes:
1. Signal-aware contexts (bd-rtp):
- Added rootCtx/rootCancel in main.go using signal.NotifyContext()
- Set up in PersistentPreRun, cancelled in PersistentPostRun
- Daemon uses same pattern in runDaemonLoop()
- Handles SIGINT/SIGTERM for graceful shutdown
2. Context propagation (bd-yb8):
- All commands now use rootCtx instead of context.Background()
- sqlite.New() receives context for cancellable operations
- Database operations respect context cancellation
- Storage layer propagates context through all queries
3. Cancellation tests (bd-2o2):
- Added import_cancellation_test.go with comprehensive tests
- Added export cancellation test in export_test.go
- Tests verify database integrity after cancellation
- All cancellation tests passing
Fixes applied during review:
- Fixed rootCtx lifecycle (removed premature defer from PersistentPreRun)
- Fixed test context contamination (reset rootCtx in test cleanup)
- Fixed export tests missing context setup
Impact:
- Pressing Ctrl+C during import/export now cancels gracefully
- No database corruption or hanging transactions
- Clean shutdown of all operations
Tested:
- go build ./cmd/bd ✓
- go test ./cmd/bd -run TestImportCancellation ✓
- go test ./cmd/bd -run TestExportCommand ✓
- Manual Ctrl+C testing verified
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Fixes GH-234 by providing automatic resolution for duplicate external_ref
values instead of forcing manual JSONL editing.
Changes:
- Add ClearDuplicateExternalRefs option to importer.Options
- Modify validateNoDuplicateExternalRefs to clear duplicates when enabled
- Keep first occurrence, clear rest when flag is set
- Enhanced error message to suggest the flag
- Add comprehensive tests for the new behavior
Usage: bd import -i issues.jsonl --clear-duplicate-external-refs
Amp-Thread-ID: https://ampcode.com/threads/T-932dcf45-76f2-4994-9b5c-a6eb20a86036
Co-authored-by: Amp <amp@ampcode.com>
When import finds same content hash with different IDs, treat it as
an update to the existing issue instead of failing with 'rename
collision' error. This handles edge cases like test data, legacy
data, or data corruption gracefully.
Amp-Thread-ID: https://ampcode.com/threads/T-e58a11be-cbbb-4a75-86d5-fc51af8f51d2
Co-authored-by: Amp <amp@ampcode.com>
**Problem:**
Closed issues were silently reopening during git sync/import operations.
When importing an issue update, the importer built an updates map with
status='closed' but NO closed_at timestamp. The UpdateIssue() function's
manageClosedAt() would only set closed_at when status was CHANGING to
closed, not when it was already closed. Result: closed_at got cleared,
effectively reopening issues.
**Root Cause:**
1. Importer built updates map without closed_at field (lines 443-451, 519-528)
2. closed_at was not in allowedUpdateFields whitelist
3. manageClosedAt() only managed closed_at for status TRANSITIONS
4. Import of already-closed issue → closed_at lost → issue reopens
**Impact:**
- WASM issues (bd-44d0, bd-8507, etc.) were closed on Nov 4 (commit 0df9144)
- They reopened as 'open' status during sync on Nov 5 (commit 8c9814a)
- Users had to repeatedly close the same issues
- Data integrity violation: status=closed with closed_at=NULL
**Fix:**
1. Add closed_at to allowedUpdateFields whitelist
2. Add closed_at to importer updates maps (both external_ref and ID paths)
3. Update manageClosedAt() to skip auto-management if closed_at explicitly provided
- Preserves import timestamps while maintaining auto-management for CLI operations
**Testing:**
- All internal/importer tests pass
- All internal/storage/sqlite tests pass
- Explicitly tests timestamp preservation in TestImportWithExternalRef
**Files Changed:**
- internal/importer/importer.go: Add closed_at to updates maps
- internal/storage/sqlite/sqlite.go: Allow closed_at updates, respect explicit values
Amp-Thread-ID: https://ampcode.com/threads/T-53ed6e45-9d04-4a35-97e9-d1ec36321ab0
Co-authored-by: Amp <amp@ampcode.com>
- Added OrphanHandling type to sqlite package with 4 modes: strict/resurrect/skip/allow
- Updated EnsureIDs() to accept orphanHandling parameter and implement mode logic
- Added CreateIssuesWithOptions() that passes orphan handling through batch creation
- Made importer.OrphanHandling an alias to sqlite.OrphanHandling
- Importer now respects opts.OrphanHandling during batch issue creation
Next: Add import.orphan_handling config and wire through CLI commands
Amp-Thread-ID: https://ampcode.com/threads/T-bb7ffdd9-f444-4975-b5f7-bfff97cb92ff
Co-authored-by: Amp <amp@ampcode.com>
- Add OrphanHandling enum: strict/resurrect/skip/allow
- Add OrphanHandling field to importer.Options
- Default to 'allow' mode to work around existing system bugs
- Strict mode can be enabled via config for safer imports
Related: bd-8072, bd-b92a
- Add sort.go with depth-based utilities (GetHierarchyDepth, SortByDepth, GroupByDepth)
- Sort issues by hierarchy depth before batch creation
- Create in depth-order batches (0→1→2→3)
- Fixes latent bug: parent-child pairs in same batch could fail if wrong order
- Comprehensive tests for all sorting functions
- Closes bd-37dd, bd-3433, bd-8b65
Part of bd-d19a (Fix import failure on missing parent issues)
Amp-Thread-ID: https://ampcode.com/threads/T-44a36985-b59c-426f-834c-60a0faa0f9fb
Co-authored-by: Amp <amp@ampcode.com>
- Add migration for UNIQUE index on external_ref column (bd-897a)
- Add validation for duplicate external_ref in batch imports (bd-7315)
- Add query planner test to verify index usage (bd-f9a1)
- Add concurrent import tests for external_ref (bd-3f6a)
The migration detects existing duplicates and fails gracefully.
Batch imports now reject duplicates with clear error messages.
Tests verify the index is actually used by SQLite query planner.
Amp-Thread-ID: https://ampcode.com/threads/T-45ca66ed-3912-46c4-963c-caa7724a9a2f
Co-authored-by: Amp <amp@ampcode.com>
- Add GetIssueByExternalRef() query function to storage interface and implementations
- Update DetectCollisions() to prioritize external_ref matching over ID matching
- Modify upsertIssues() to handle external_ref matches in import logic
- Add index on external_ref column for performance
- Add comprehensive tests for external_ref matching in both collision detection and import
- Enables re-syncing from external systems (Jira, GitHub, Linear) without duplicates
- Preserves local issues (no external_ref) from being overwritten
- Always recompute content_hash in importer to avoid stale hashes from JSONL
- Add .gitignore to test repos to prevent database files from being tracked
- Fix daemon auto-import test to use correct RPC operation ('show' not 'get_issue')
- Set last_import_time metadata in test helper to enable staleness check
- Add filesystem settle delay after git pull in tests
Root cause: CloseIssue updates status but not content_hash, so importer
thought issues were unchanged. Always recomputing content_hash fixes this.
Amp-Thread-ID: https://ampcode.com/threads/T-63ef3a7d-8efe-472d-97ed-6ac95bd8318b
Co-authored-by: Amp <amp@ampcode.com>
- Fix Windows binary path issues (bd.exe vs bd)
- Skip scripttest on Windows (requires Unix shell)
- Skip file lock tests on Windows (platform locking differences)
- Fix registry tests to use USERPROFILE on Windows
- Fix 8 unparam lint warnings by marking unused params with _
All changes are platform-aware and maintain functionality.
Amp-Thread-ID: https://ampcode.com/threads/T-bc27021a-65db-4b64-a3f3-4e8d7bc8aa0d
Co-authored-by: Amp <amp@ampcode.com>
