When users delete local BME IndexedDB via UI actions (delete current/all
IDB), the chat metadata's st_bme_commit_marker was not cleared. This left
an accepted high-revision promise with no local DB backing, causing
persist-mismatch:indexeddb-behind-commit-marker and blocking graph load
indefinitely.
- index.js: add clearCurrentChatCommitMarker() helper and expose via runtime
- ui-actions-controller.js: call clearCurrentChatCommitMarker before
syncGraphLoadFromLiveContext after IDB deletion
- p0-regressions.mjs: regression test asserting marker is cleared before
reload after current-IDB deletion
- sync/bme-sync.js: conservatively clear backend hash mappings and mark
vectorIndexState dirty before importing remote snapshots via download,
merge, and cloud backup restore, preventing stale clean-looking state
after cross-device sync or restore
- vector/vector-index.js: mark backend vector state dirty on real backend
query failures (HTTP/network) instead of silently returning empty results
- regression: indexeddb-sync.mjs covers download/restore/merge import
dirty marking; p0-regressions.mjs covers backend query failure dirtying
- reconcileIndexedDbProbeFailureState now handles persist-mismatch:indexeddb-* reasons, ensuring bounded retries and transition to BLOCKED instead of hanging forever in LOADING
- expose clearCachedIndexedDbSnapshot, clearAllCachedIndexedDbSnapshots, closeBmeDb, closeAllBmeDbs, syncGraphLoadFromLiveContext in runtime context for UI actions
- onDeleteCurrentIdbController / onDeleteAllIdbController now clear in-memory caches, close DB handles, and force a fresh graph load sync after deletion
- humanize persist-mismatch:indexeddb-behind-commit-marker panel messages into user-friendly Chinese text
- add regression test for empty IndexedDB + accepted commit marker mismatch fallback-to-blocked scenario
- Add hasContiguousJournalCoverageThroughFloor to check if current
batchJournal contiguously covers back through the old retained floor
- Add reconcileManualBackupBatchJournalCoverage that clears stale
coverage when actual journals have bridged the gap
- Self-clean during normalizeGraphRuntimeState (load-time) and
appendBatchJournal (write-time)
- findJournalRecoveryPoint now uses reconciled coverage so stale
manual backup floors no longer block valid reverse-journal recovery
- Add regression tests for gap-not-yet-bridged and bridged scenarios
- prompt-builder.js: add RECALL_TARGET_CONTENT_HEADER, update splitSectionedTranscriptPayloadMessage to recognize recall-specific target header
- retriever.js: add buildRecallSectionedTranscript helper, format recentMessages as sectioned transcript with context-review and recall-target headers for prompt building while keeping flat string[] for ranking
- p0-regressions.mjs: add testRecallUsesSectionedPromptMessagesForContextAndTarget regression asserting two system messages with correct transcriptSection and headers
