This release includes breaking changes for platform teams planning a safe upgrade.
✓ No known CVEs patched in this version
Summary
AI summaryFixes indexer hangs, unifies ignore rules, and enables partial search results during indexing.
Full changelog
v0.4.6 — Indexer Hang Fix + Unified Ignores + Partial Search
Fix A: Fail-fast / Status Recovery
IndexState::Failednow actually gets set — previously defined but never written, causing infinite hang at ~99% indexingfailed_files: Vec<String>tracks per-file read errors with diagnosticsfailed_embeddings: u32distinguishes embedding failures from file-read failuresFailedstate auto-triggers re-index on nextindex_projectrequest- Stale
EmbeddingPendingdetection (60s) — completion monitor forces transition if embeddings stop progressing - Stale
Indexingdetection (300s) — completion monitor catches stuck indexing pipeline and setsFailed
Fix B: Unified Ignore System
- Single
is_ignored_file()function inscanner.rswithOverrideBuilder— one source of truth - Watcher and indexer both delegate to scanner (removed duplicate hardcoded lists)
- Flutter/Dart defaults added:
.g.dart,.freezed.dart,.arb,build/,*.mocks.dart
Fix C: Partial Search During Indexing
search_codenow returns partial results withis_partial: trueflag during indexing (previously hard-blocked with zero results)recall_codehas matchingis_partialflag for consistency
Internal
- SurrealDB schema updated with
failed_embeddingsfield definition - Progress logging now includes failed file count
Full Changelog: https://github.com/pomazanbohdan/memory-mcp-1file/compare/v0.4.5...v0.4.6
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
Share this release
About pomazanbohdan/memory-mcp-1file
A self-contained Memory server with single-binary architecture (embedded DB & models, no dependencies). Provides persistent semantic and graph-based memory for AI agents.
Related context
Beta — feedback welcome: [email protected]