Mnemon Mcp
Gives Claude persistent memory across sessions using a local SQLite database with FTS5 search and optional vector similarity. Organizes memories into four layers (episodic, semantic, procedural, resource) with different decay rates and access patterns. Ten tools handle adding, searching, updating, and versioning facts. Facts form superseding chains instead of being overwritten, so you can trace how knowledge evolved. Search combines BM25 ranking with recency and importance weighting, plus hybrid mode when you bring your own embeddings via OpenAI or Ollama. Includes session tracking, health diagnostics, and export to JSON or Markdown. Reach for this when you want your agent to remember user preferences, project conventions, or past conversations without sending anything to the cloud.
- GitHub stars
- 3
- tools
- 10
- auth
- Secret needed
- transport
- STDIO
- env vars
- 3
- last updated
- Mar 16, 2026
Tools
10 toolsPublic tool metadata for what this MCP can expose to an agent.
memory_addStore a memory with layer, entity, confidence, importance, and optional TTL
Store a memory with layer, entity, confidence, importance, and optional TTL
No parameter schema in public metadata yet.
memory_searchFull-text or exact search with filters by layer, entity, date, scope, confidence
Full-text or exact search with filters by layer, entity, date, scope, confidence
No parameter schema in public metadata yet.
memory_updateUpdate in-place or create a versioned replacement (superseding chain)
Update in-place or create a versioned replacement (superseding chain)
No parameter schema in public metadata yet.
memory_deleteDelete a memory; re-activates its predecessor if any
Delete a memory; re-activates its predecessor if any
No parameter schema in public metadata yet.
memory_inspectGet layer statistics or trace a single memory's version history
Get layer statistics or trace a single memory's version history
No parameter schema in public metadata yet.
memory_exportExport to JSON, Markdown, or Claude-md format with filters
Export to JSON, Markdown, or Claude-md format with filters
No parameter schema in public metadata yet.
memory_healthRun diagnostics: expired entries, orphaned chains, stale memories; optionally GC
Run diagnostics: expired entries, orphaned chains, stale memories; optionally GC
No parameter schema in public metadata yet.
memory_session_startStart an agent session — returns session ID for grouping memories
Start an agent session — returns session ID for grouping memories
No parameter schema in public metadata yet.
memory_session_endEnd a session with optional summary; returns duration and memory count
End a session with optional summary; returns duration and memory count
No parameter schema in public metadata yet.
memory_session_listList sessions with filters by client, project, or active status
List sessions with filters by client, project, or active status
No parameter schema in public metadata yet.
Sponsors
Agent CaffeineSponsoredClose your lid, come back to finished work
SnugSponsoredThe clipboard manager for the AI eraPromote your product hereReach 125k+ visitors monthly
WaxSponsoredBlock anything with supervision, without erasing your iPhonePromote your product hereReach 125k+ visitors monthly
knobsSponsoredSee your site on any phone, any day, in any languageConfiguration
MNEMON_DB_PATHPath to SQLite database file (default: ~/.mnemon-mcp/memory.db)
MNEMON_EMBEDDING_PROVIDEREmbedding provider for vector search: 'openai' or 'ollama' (default: disabled)
MNEMON_EMBEDDING_API_KEYsecretAPI key for embedding provider (required for OpenAI)
Readme
mnemon-mcp
Persistent layered memory for AI agents. Local-first. Zero-cloud. Single SQLite file.
Landing Page · npm · GitHub
Your AI agent forgets everything after each session. Mnemon fixes that.
It gives any MCP-compatible client — OpenClaw, Claude Code, Cursor, Windsurf, or your own — a structured long-term memory backed by a single SQLite database on your machine. No API keys, no cloud, no telemetry. Just npm install and your agent remembers.
Details
- Repository
- github.com/nikitacometa/mnemon-mcp
- Categories
- Databases, AI & LLM Tools
- Package
mnemon-mcp- Transport
- STDIO
- Auth
- Required
- Registry
- MCP registry, active
Related Databases MCP servers
View all
A2dbagentic-eng/a2dbSafe, read-only database access for AI agents. 5 databases, batch queries, SQLGlot.2stars
Virtual Filesystemlu-zhengda/mcp-virtual-fsPostgreSQL-backed virtual filesystem for agents: session isolation, shared stores, POSIX ops, RLS.2stars
Shrikeshrike-security/shrike-mcpGoverns what AI agents do: tool calls, SQL, commands, files checked against policy before they run.2stars
ClickHousealyiox/mcp-clickhousexRead-only MCP server for ClickHouse metadata, parameterized queries, and plan analysis.1stars
Fastbcp Mcparpe-io/fastbcp-mcpMCP server for FastBCP — high-performance parallel database export to files and cloud1starsLakexpress Mcparpe-io/lakexpress-mcpMCP server for LakeXpress — automated database-to-cloud data pipeline as Parquet1stars