openclaw

mirror of https://github.com/openclaw/openclaw.git synced 2026-05-30 08:55:04 +00:00

Files

Tarun Sukhani fff48a146d memory-neo4j: add auto-recall filtering, assistant capture, importance scoring, conflict detection

Five high-impact improvements to the memory system:

1. Min RRF score threshold on auto-recall (default 0.25) — filters low-relevance
   results before injecting into context
2. Deduplicate auto-recall against core memories already present in context
3. Capture assistant messages (decisions, recommendations, synthesized facts)
   with stricter attention gating and "auto-capture-assistant" source type
4. LLM-judged importance scoring at capture time (0.1-1.0) with 5s timeout
   fallback to 0.5, replacing the flat 0.5 default
5. Conflict detection in sleep cycle (Phase 1b) — finds contradictory memories
   sharing entities, uses LLM to resolve, invalidates the loser

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

2026-02-16 17:56:38 +08:00

config.test.ts

memory-neo4j: add auto-recall filtering, assistant capture, importance scoring, conflict detection

2026-02-16 17:56:38 +08:00

config.ts

memory-neo4j: add auto-recall filtering, assistant capture, importance scoring, conflict detection

2026-02-16 17:56:38 +08:00

embeddings.test.ts

memory-neo4j: add context-length-aware embedding truncation

2026-02-16 17:56:38 +08:00

embeddings.ts

memory-neo4j: strip injected context blocks, add core category, widen embeddings context