{"$schema":"https://wellknown.network/schemas/agent-record-v1.json","schemaVersion":"1","id":"ag_hg8hmky3jhhe","handle":"mnestic-mcp","url":"https://wellknown.network/agents/mnestic-mcp","links":{"self":"https://wellknown.network/agents/mnestic-mcp/record.json","html":"https://wellknown.network/agents/mnestic-mcp","markdown":"https://wellknown.network/agents/mnestic-mcp/record.md","api":"https://wellknown.network/api/v1/agents/mnestic-mcp","status":"https://wellknown.network/api/v1/agents/mnestic-mcp/status","claim":"https://wellknown.network/agents/mnestic-mcp/claim","claimApi":"https://wellknown.network/api/v1/claims","claimDescriptor":"https://wellknown.network/agents/mnestic-mcp/claim.json","badge":"https://wellknown.network/agents/mnestic-mcp/badge.svg","openapi":"https://wellknown.network/openapi.json","history":"https://wellknown.network/api/v1/agents/mnestic-mcp/history","tools":"https://wellknown.network/api/v1/agents/mnestic-mcp/tools"},"ard":{"identifier":"urn:air::server:mnestic-mcp","type":"application/mcp-server-card+json"},"kind":"mcp_server","declared":{"name":"mnestic-mcp","summary":"Local memory MCP server backed by mnestic — hybrid BM25+vector+graph recall with per-leg attribution and time travel, all on your machine.","description":"# mnestic-mcp\n\nA local **memory MCP server** backed by\n[mnestic](https://github.com/shuruheel/mnestic) — an embedded graph + vector +\nfull-text database (a maintained fork of CozoDB). One process, one file, fully\nlocal: your agent's memory never leaves the machine.\n\n> mnestic is a maintained fork of [CozoDB](https://github.com/cozodb/cozo); it is not the official CozoDB. Original design credit belongs to Ziyang Hu and the Cozo Project Authors.\n\n## Install\n\nAdd to any MCP client (Claude Desktop, Claude Code, Cursor, ...):\n\n```json\n{ \"command\": \"uvx\", \"args\": [\"mnestic-mcp\"] }\n```\n\nThat's it. The database lives in your platform data dir (override with\n`--db PATH` or `MNESTIC_MCP_DB`); embeddings run locally via\n[fastembed](https://github.com/qdrant/fastembed) (default\n`BAAI/bge-small-en-v1.5`, a one-time ~67 MB download that happens in the\nbackground — keyword search works immediately, before the model arrives).\nPre-provision with `uvx mnestic-mcp --download-model`.\n\n## Tools\n\n`store`, `store_batch`, `search`, `find_related`, `list_recent`, `update`,\n`delete`, `link`, `recall_as_of`, `stats` — plus the two things no other local\nmemory server has:\n\n- **`search(explain=true)`** — per-leg attribution: exactly how much the\n  keyword (BM25), vector (HNSW), and graph-proximity legs each contributed to\n  every result, straight from the engine's fused three-way retrieval. Ask\n  your agent *\"why did you recall that?\"* and get a real answer.\n- **`recall_as_of(t)`** — time travel. Updates and deletes are never\n  destructive (valid-time history in the engine's bitemporal storage):\n  *\"what did you know about this before last Tuesday?\"* just works.\n\n`search` is keyword-first with an automatic hybrid fallback; `link` builds a\ntyped, weighted memory graph that both `find_related` (budget-bounded\ntraversal) and the hybrid graph leg exploit.\n\n**Byte-budgeted output** (0.2.0): every wide-result tool (`search`,\n`find_related`, `list_recent`, `recall_as_of`) takes `max_bytes` (default\n2…","publisher":{"name":"Shan Rizvi","url":null},"homepage":"https://github.com/shuruheel/mnestic","repository":"https://github.com/shuruheel/mnestic","version":"0.2.0","license":"MPL-2.0","protocols":["mcp"],"tags":["agent-memory","cozodb","hybrid-search","mcp","memory","mnestic","model-context-protocol"],"pricing":null,"endpoints":[{"url":"pypi:mnestic-mcp","type":"package_pypi","auth":null,"probeable":false}],"skills":null,"tools":null,"extra":null,"attribution":{"kind":"pypi","name":"pypi","license":"pypi","repoUrl":"pypi","summary":"pypi","version":"pypi","description":"pypi","homepageUrl":"pypi","publisherName":"pypi"}},"derived":{"capabilities":[{"slug":"data.vector-search","name":"Vector Search","confidence":1,"provenance":"derived"},{"slug":"knowledge.memory","name":"Agent Memory","confidence":1,"provenance":"declared"}],"categories":["data","knowledge"],"language":"en"},"observed":{"status":"unknown","statusReason":"Distributed as a package to run locally; no network endpoint to check.","lastOkAt":null,"lastProbedAt":null,"statusComputedAt":null,"reliability30d":null,"latestObservations":[],"tools":null,"package":{"name":"mnestic-mcp","registry":"pypi","observedAt":"2026-09-10T06:25:19.109Z","publishedAt":"2026-08-08T21:56:54.821717Z","latestVersion":"0.2.0"},"toolSurface":null,"endpointFacts":[]},"verification":{"claimed":false,"claimedAt":null,"proofs":[]},"provenance":{"sources":[{"source":"pypi","key":"mnestic-mcp","url":"https://pypi.org/project/mnestic-mcp/","firstSeenAt":"2026-09-10T06:23:29.270Z","fetchedAt":"2026-09-10T06:23:29.270Z","normalizedAt":"2026-09-10T06:23:29.270Z"}]},"firstSeenAt":"2026-09-10T06:23:29.270Z","updatedAt":"2026-09-10T06:25:19.109Z"}