v10.47 — Now Available
MCP Memory Service
Persistent memory for AI agents. Semantic search, knowledge graph, automatic consolidation — now accessible from Claude.ai via Streamable HTTP.
v10.47 — Now Available
Persistent memory for AI agents. Semantic search, knowledge graph, automatic consolidation — now accessible from Claude.ai via Streamable HTTP.
One-call maintenance orchestrator + pre-quantized DeBERTa Docker images. ~1,780 tests.
Call memory_quality(action='maintain') to run a full maintenance cycle — cleanup, conflict detection, stale detection, and quality snapshot — in one shot. dry_run=true by default: always safe to explore. (PR #802, @filhocf, closes #799)
Opt-in via MCP_MAINTAIN_AUTO_RESOLVE. Pairs with similarity ≥ 0.95, same memory type, and age delta > 7 days are resolved automatically — newer memory wins. Threshold and age are configurable.
DeBERTa ONNX classifier is now quantized (fp16 / int8) at Docker build time. :quality-cpu image drops from ~1.7 GB to ~600 MB. Cold starts no longer re-run quantization. (PR #803, closes #793)
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.