Advertise
MCP.so
Sign In

Mnemon — Persistent Layered Memory for AI Agents MCP Server

@nikitacometa

Repository

Basic information

Category

Memory & Knowledge

Transports

stdio

Author

nikitacometa

Submitted by

Nikita

About Mnemon — Persistent Layered Memory for AI Agents MCP Server

Persistent 4-layer memory (episodic, semantic, procedural, resource) backed by SQLite FTS5. Fact versioning, Snowball stemming (EN+RU), BM25 ranking. Zero-cloud, single-file database. 7 MCP tools.

#memory#sqlite#persistent-memory#ai-agent#fts5#multilingual#knowledge-base

Mnemon — Persistent Layered Memory for AI Agents MCP Server config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "mnemon-mcp": {
      "command": "mnemon-mcp"
    }
  }
}

Mnemon — Persistent Layered Memory for AI Agents MCP Server tools

Tool list not available yet

This listing does not currently have a tool list to display. This does not mean the server has no tools; check its documentation for supported capabilities.

We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.

Overview

What is Mnemon — Persistent Layered Memory for AI Agents?

Mnemon — Persistent Layered Memory for AI Agents gives any MCP-compatible client (Claude Code, Cursor, Windsurf, or your own) a structured long-term memory backed by a single SQLite database on your machine. It uses a layered memory model (episodic, semantic, procedural, resource) with different lifetimes and access patterns, all without API keys, cloud services, or telemetry.

How to use Mnemon — Persistent Layered Memory for AI Agents?

Install globally via npm: npm install -g mnemon-mcp. Then configure the server in your MCP client’s configuration file. The server exposes seven MCP tools for storing, searching, updating, deleting, inspecting, exporting, and maintaining memory entries.

Key features of Mnemon — Persistent Layered Memory for AI Agents

  • Four memory layers with configurable lifetimes (decay, stable, rarely changes)
  • Seven MCP tools: add, search, update, delete, inspect, export, health
  • Full-text search (FTS5) with BM25 ranking and Snowball stemming
  • Optional vector search via OpenAI or Ollama embeddings with hybrid ranking
  • Fact versioning: version chains with superseding, full history via memory_inspect
  • Fully local: single SQLite database, no external services or API keys

Use cases of Mnemon — Persistent Layered Memory for AI Agents

  • Give AI agents persistent long-term memory across sessions
  • Store and retrieve user preferences, facts, and relationships
  • Maintain workflow rules, conventions, and procedural knowledge
  • Keep reference material and book notes with automatic decay
  • Log events and sessions for episodic recall

Frequently asked questions

1

What memory layers does Mnemon provide?

Episodic (events, sessions; decays with 30-day half-life), Semantic (facts, preferences; stable), Procedural (rules, workflows; rarely changes), and Resource (reference material; decays slowly over 90 days).

2

Does Mnemon require any cloud services or API keys?

No. It runs entirely locally with a single SQLite database. No external services, no API keys, no telemetry.

3

How does search work in Mnemon?

Search uses FTS5 with BM25 ranking, multi-word AND with progressive OR fallback, and Snowball stemmer for English and Russian. Optional vector search via OpenAI or Ollama embeddings uses hybrid ranking with Reciprocal Rank Fusion.

4

What are the system requirements for Mnemon?

Node.js ≥22. No other external dependencies.

5

How does Mnemon handle fact updates?

Updates can be performed in-place or as versioned replacements (superseding chain). Search always returns the latest version, while memory_inspect reveals the full version history. Deleting a memory reactivates its predecessor if any.

Comments

More Memory & Knowledge MCP servers

AOCI-CODE

aoci-spec

Local MCP server and CLI that keeps a Git-versioned, model-written index of a codebase and its database schema, so coding agents start each task with the whole system in view.

1.7K

Agimem

Persistent memory for AI agents. Hosted MCP server with capsule-scoped key/value storage, tags, TTL expiration, bulk ops, and RBAC.

server-noonien - shared memory MCP

sequico

Shared, serverless, conflict-free memory for AI agents: a multi-machine drop-in for the MCP memory knowledge graph.

Azquo

Azquo Holdings Ltd

Exact, shared memory for AI agents: a hierarchical database where totals always add up, every write is attributed to the agent and person behind it, and any figure traces back to its source file. Agents create databases,

Neither MCP

stonianua

Project context your AI can query through MCP — with source evidence.

Advertise