github.com/orneryd/Mimir

Added

• Added CHANGELOG.md at repository root for release history and guidance.

Changed

• src/api/chat-api.ts: Use database-aware default for semantic similarity (NornicDB = 0.5, Neo4j = configured default) so server-side embeddings use the correct threshold.
• testing/file-indexer-nornicdb.test.ts: Mock improvements for EmbeddingsService exports and corrected expectations to match NornicDB behavior.

Fixed

• Unit tests: fixed failing mocks and assertions; all relevant unit tests pass locally (including file-indexer-nornicdb tests).

Misc

• Small tooling/test cleanup and formatting tweaks made while validating the changes.

v1.0.1: additional nornicDB enhancements

Added

• NornicDB server-side hybrid search support: pass string queries to db.index.vector.queryNodes so the server generates embeddings.

Changed

• Fixed NornicDB fulltext/Hybrid LIMIT handling so LIMIT is respected by server-side queries.
• Standardized similarity semantics: both NornicDB and Neo4j return cosine similarity in the 0.0–1.0 range.
• Set default cosine similarity thresholds: NornicDB default min similarity = 0.5, Neo4j default = 0.75.
• Removed special-case quirk-handling code paths for NornicDB; unified search paths and updated inline comments.

Fixed

• Tests and tooling:
  • Unit tests covering UnifiedSearchService (mocked) pass (19 tests).
  • Live integration tests (NornicDB) pass when enabled; these are opt-in via RUN_LIVE_TESTS=true (15 tests).

Performance

• k-means clustering now runs on a 15-minute timer rather than on-trigger to avoid blocking operations.

Files touched (high-level)

• src/managers/UnifiedSearchService.ts — NornicDB hybrid search, threshold and comments
• src/api/chat-api.ts — use NornicDB-specific threshold when initiating semantic searches
• src/managers/GraphManager.ts — comment cleanup
• src/types/IGraphManager.ts — comment cleanup
• testing/nornicdb-live-integration.test.ts — skip-by-default with RUN_LIVE_TESTS

Tagged as v1.0.1 (annotated tag): "v1.0.1: additional nornicDB enhancements"

For full details, see the commits between v1.0.0 and v1.0.1.

Status: Production-ready baseline - Preserved in v2.0.0

Proven Performance:

• Score: 82/100 (Tier A)
• Created actual test file
• Showed evidence
• Correct file scope
• Precise analysis

What Makes It Work (DO NOT CHANGE):

1. ✅ Clear file scope prevents wrong edits
2. ✅ Evidence requirements push toward action
3. ✅ Chain-of-thought template provides structure
4. ✅ Complete example shows what "done" looks like
5. ✅ Phase-based workflow with clear steps
6. ✅ Anti-patterns explicitly listed

Known Gaps (Can improve):

• Skipped baseline tests (-5 pts)
• No debug markers (-5 pts)
• Stopped after 1 bug (needs 3+)
• Asked for direction (needs autonomy)

Target for next version: 90+ (Tier S) by addressing gaps WITHOUT breaking what works

Release v0.1.0 - async lambdas