Skip to content
MCP server

Mnemosyne OS

Local, human-governed memory vaults that coding agents can read through MCP

Runs locallyStatus and history
Install
Pricing
Freemium
Platforms
macOS, Windows, Linux
Runs
Locally
Version
2.4.0
Listing updated
2 days ago

About Mnemosyne OS

Mnemosyne OS is a freemium MCP server for knowledge and memory. It runs locally on your machine and is started by your AI client.

Mnemosyne OS is a desktop app that keeps your AI memory on your own machine, and this npm package is the door that lets agents such as Claude Code, Cursor or Claude Desktop reach it. Memory is organised into vaults, user-defined domains like code, notes or a journal; the server only reaches the vaults you list in MNEMO_VAULTS, with a default vault used when a call does not name one.

The design stresses control. The site says every program that reads your memory identifies itself first, and that nothing becomes permanent without your hand, so no agent can write a lasting memory from the network. The registry entry also describes access to what other coding agents on the same machine wrote. Agents can be pointed at folders such as .claude, .cursor or .aider to index their own notes.

The wider product adds a chat that cites your documents, a graph view of memory, a document reader, voice, an image studio and small apps called cartridges. The site says there is no account and no mandatory server, a local model can run offline, and you bring your own API key or buy credits.

Install

Run in a terminal

claude mcp add mcp -- npx -y @mnemosyne_os/mcp

Add to claude_desktop_config.json

{
  "mcpServers": {
    "mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@mnemosyne_os/mcp"
      ]
    }
  }
}
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@mnemosyne_os/mcp"
      ]
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "mcp": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@mnemosyne_os/mcp"
      ]
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@mnemosyne_os/mcp"
      ]
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@mnemosyne_os/mcp"
      ]
    }
  }
}

Configuration

  • MNEMO_DEFAULT_VAULTVault used when a call does not name one, and always reachable. Vaults are the user's own domains of memory, such as code, notes or a journal. Left unset, the server uses the first vault of MNEMO_VAULTS, or DEV when there is no list.
  • MNEMO_VAULTSComma-separated list of the vaults this server may reach, besides the default. Once set, it is the whole list: a vault left out is refused. Left unset, the server reaches DEV, PERSONAL and SOCIAL.
  • MNEMO_SDK_PORTPort of the local Mnemosyne OS SDK server the desktop app listens on.

What it does

Vault-scoped memory
Agents reach only the vaults you list.
Identified access
Each reader identifies itself, even locally.
Human-approved writes
No agent can create a permanent memory from the network.
Shared agent notes
Index folders like .claude, .cursor or .aider.
Local-first
Runs on your machine with no account; cloud is optional.
Wider app
Cited chat, memory graph, voice and document reader.

Good for

  • Give a coding agent persistent memory across sessions
  • Keep personal notes private yet queryable by agents
  • Share what one coding agent learned with another
  • Question your own documents offline with a local model

Pricing

Freemium

No subscription; bring your own API key or buy credits

Prices change. Check mnemosyne-os.io before you buy.

Mnemosyne OS FAQ

Does it need an account?

No. The site says no signup, password or mandatory server; you download and launch.

How do I connect an agent?

Run npx -y @mnemosyne_os/mcp to connect Claude Code, Cursor or Claude Desktop to your vaults.

Can agents write memories?

The site says nothing is kept without you and no agent can write one from the network.

Does it need internet?

The engine runs on your machine; with a local model you can work offline.

Is Mnemosyne OS free?

Mnemosyne OS has a free plan, with paid plans for more usage or features.

How do I add Mnemosyne OS to Claude Code?

Run this command in your terminal: claude mcp add mcp -- npx -y @mnemosyne_os/mcp

What are the best alternatives to Mnemosyne OS?

The closest alternatives to Mnemosyne OS are Codebase Memory, XERJ, Claude Code Ultimate Guide, Nautilus Compass Memory and note1.

Alternatives to Mnemosyne OS

  1. Codebase Memory

    MCP server

    Local MCP server that builds a code knowledge graph for AI coding agents

    Runs locally
    • Open source
    • 48.7K GitHub stars
    • Knowledge and memory
  2. XERJ

    MCP server

    Local search engine for AI agents with cited answers over your code and docs

    Runs locally
    • Open source
    • 19.2K GitHub stars
    • Knowledge and memory
  3. MCP server for searching the Claude Code Ultimate Guide from any MCP client

    Runs locally
    • Open source
    • 6.5K GitHub stars
    • AI models and LLM tooling
  4. Local-first long-term memory layer for AI agents, served over MCP

    Not checked yet
    • 5 tools
    • Remote
    • Open source
    • 1.3K GitHub stars
    • Knowledge and memory
  5. note1

    MCP server

    MCP server giving AI assistants search and summaries across your note1 meetings

    • 4 tools
    • Remote
    • Free trial
    • 1 GitHub star
    • Productivity