Enables AI agents and MCP-compatible clients to perform hybrid episodic long-term memory retrieval by combining BM25 keyword search, semantic n-gram cosine similarity, and recency decay ranking over stored memories.
Enables agents to retrieve relevant context via a zero-dependency hybrid search engine that blends sparse keyword-based lexical ranking with dense vector similarity scores into one fused result set, then refines it with knowledge graph triplet expansion and lost-in-the-middle context reordering. It also provides episodic memory consolidation, recency decay, and semantic query caching, exposed as a JSON-RPC 2.0 stdio MCP server for clients such as Claude Desktop, Cursor, and Windsurf.
Local deterministic BM25 memory for AI agents — offline-first, no API key, SHA-256 content-addressed shards, stdio MCP transport. Same query always returns the same ranked result.
MCP-native persistent memory layer for AI agents across Claude Code, Cursor, VS Code, and OpenClaw. Powered by hybrid vector search, BM25, and cross-encoder reranking with a published 73.1% LoCoMo benchmark accuracy.
Enables deterministic, zero-dependency long-horizon conversational memory compaction and episodic anchor extraction for AI agents, with native MCP protocol support and structured JSON telemetry output.
Provides persistent, searchable memory with hybrid keyword and semantic search, storing memories in a single SQLite file without external dependencies.
