[Dream Cycle 2026-06-08] memory: ADR-147 multi-signal retrieval (semantic+BM25+entity)#2317
Draft
ruvnet wants to merge 1 commit into
Draft
[Dream Cycle 2026-06-08] memory: ADR-147 multi-signal retrieval (semantic+BM25+entity)#2317ruvnet wants to merge 1 commit into
ruvnet wants to merge 1 commit into
Conversation
… Mem0 SOTA (94.4% LongMemEval) ADR-147 proposes adding parallel semantic+BM25/FTS5+entity retrieval with RRF fusion to UnifiedMemoryService. FTS5 infrastructure already in codebase (fts5.ts, graceful-retrieval.ts); architectural decision records the retrieval contract change and 4-phase implementation plan. https://claude.ai/code/session_01R4sETrdwMxLBZCWXowsagM
This was referenced Jun 8, 2026
Owner
Author
ruvnet
added a commit
that referenced
this pull request
Jun 8, 2026
…2327) Adds entity matching as a third RRF arm alongside dense + sparse, plus per-result signal provenance (#2317). What this lands: - `entity-tagger.ts` — regex extractor for emails, URLs, file paths, quoted phrases, proper-noun 2-grams (12 unit tests) - `hybridSearch` controller — entity arm wired into the existing RRF+MMR fusion at `controller-registry.ts` - `signals: ('vector' | 'bm25' | 'entity')[]` on every fused result End-to-end capability smoke against built dist (not just TS source): - Corpus: 30 generic "authentication" entries + 1 "Alice Smith" needle - Query: `"Alice Smith authentication"` - Result: **Alice ranks #1** with `signals=["vector","bm25","entity"]`; runners-up only have `["vector","bm25"]` - RRF score gap: 0.0477 vs 0.0323 = **47% boost** from the entity signal 16/16 tests in touched modules. Full memory suite 416/420 (4 pre-existing Windows-env failures in unrelated files). Closes part of #2317. The ADR's stated P1 ("wire FTS5 + RRF") was already shipped before this PR; this lands the actual P2 (entity arm + provenance) that the dream-cycle author missed when surveying the tree.
ruvnet
added a commit
that referenced
this pull request
Jun 8, 2026
#2327) @claude-flow/memory 3.0.0-alpha.19 → 3.0.0-alpha.20. Adds the entity arm to hybridSearch alongside the existing dense + sparse RRF fusion, plus per-result signals: ('vector'|'bm25'|'entity')[] provenance. End-to-end capability smoke against built dist confirmed: Alice needle in 31-doc corpus ranks #1 with all three signals; runner-up has only vector+bm25 — RRF score gap of ~47%. @claude-flow/cli, claude-flow, ruflo 3.10.38 → 3.10.39. CLI also pins @claude-flow/memory to ^3.0.0-alpha.20 so the wrapper users pick up the entity arm automatically. All four packages published with latest+alpha+v3alpha aligned. Lockfile regen included (lesson from #2311 — bumping a workspace dep without regenerating v3/pnpm-lock.yaml breaks frozen-lockfile CI). Co-Authored-By: RuFlo <ruv@ruv.net>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Tonight's Rotation
d065b15927c6ba7318623e8af123e7980e4c6681What's in this PR
ADR-147 —
v3/docs/adr/ADR-147-multi-signal-memory-retrieval.mdProposes replacing Ruflo's vector-only memory retrieval with a parallel multi-signal strategy (semantic HNSW + BM25/FTS5 + entity matching) fused via Reciprocal Rank Fusion. No implementation code in this PR — architectural decision only, per dream-cycle protocol.
README row —
v3/docs/adr/README.mdupdated with ADR-147 entry.Research Summary
The 2026-06-08 dream session found that SOTA agent memory retrieval has shifted decisively from single-vector to multi-signal:
Ruflo gap: vector-only retrieval, no LoCoMo/LongMemEval scores published,
fts5.tsandgraceful-retrieval.tsexist but are unwired.ADR-147 closes the gap in 4 phases: wire FTS5 → add entity tagger → async writes by default → publish benchmark scores.
ADR Link
ADR-147 — Multi-Signal Memory Retrieval
Gist
Dream Cycle report: SHA-256
fac0b931deb16a4c41d96a1864226fb9ca363ab8018bda88602d457566f962a9Witness stamp:
419823cd839e080543d4d5ddae8e7e0d9642bbce0f9ca165c6e97574afc703d9Merge Policy
Do not self-merge. Leave for human review. This is a nightly research artifact — merge when the engineering team decides to act on ADR-147's implementation plan.
Generated by Claude Code