v10.43 — Now Available
MCP Memory Service
Persistent memory for AI agents. Semantic search, knowledge graph, automatic consolidation — now accessible from Claude.ai via Streamable HTTP.
v10.43 — Now Available
Persistent memory for AI agents. Semantic search, knowledge graph, automatic consolidation — now accessible from Claude.ai via Streamable HTTP.
Reciprocal Rank Fusion (RRF) for SQLite-vec hybrid search: industry-standard result fusion, configurable via env vars. ~1,732 tests.
Set MCP_HYBRID_FUSION_METHOD=rrf to use RRF (Cormack, Clarke & Buettcher 2009) for combining vector + keyword search results in the SQLite-vec backend. Default remains weighted_average — zero breaking changes. (PR #773, @filhocf)
Tune RRF behaviour with MCP_HYBRID_RRF_K (smoothing constant, default 60) and MCP_HYBRID_RRF_CONSENSUS_BOOST (bonus for documents ranked by both retrievers, default 0.1). 10 new tests in test_rrf_fusion.py. (PR #773)
RRF is opt-in. Existing deployments using weighted_average (or no explicit setting) are unaffected. Implementation scoped to the SQLite-vec backend, matching the hybrid search scope of issue #771. (PR #773, @filhocf)
Claude.ai connects to your memory server through Streamable HTTP with OAuth 2.1 authentication.
Battle-tested with comprehensive testing and optimized for performance.
Install from PyPI and connect your AI agents to persistent memory.