Mnemon
LLM-supervised persistent memory for AI agents — graph-based recall, cross-session knowledge, single binary. Works with Claude Code, OpenClaw, and any CLI agent.
What is Mnemon?
Mnemon is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to llm-supervised persistent memory for ai agents — graph-based recall, cross-session knowledge, single binary. works with claude code, openclaw, and any cli agent.
LLM-supervised persistent memory for AI agents — graph-based recall, cross-session knowledge, single binary. Works with Claude Code, OpenClaw, and any CLI agent.
This server falls under the Knowledge & Memory category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- LLM-supervised persistent memory for AI agents — graph-based
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx mnemonConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Mnemon
Mnemon is a single-binary, graph-based persistent memory server for AI agents that stores facts, links them with typed semantic relationships (temporal, causal, entity, semantic), and retrieves them via graph traversal and optional vector search. It is LLM-supervised — the host AI makes judgment calls about what to remember while Mnemon handles the deterministic storage and retrieval — giving agents durable cross-session knowledge without growing the context window. Written in Go and distributed as a standalone binary, it integrates with Claude Code, Codex, OpenClaw, and other CLI agents through a single setup command.
Prerequisites
- Go 1.24+ (if installing via go install or building from source)
- Homebrew (macOS/Linux) for the easiest installation via brew
- An MCP-compatible agent: Claude Code, Codex, OpenClaw, Pi, Hermes, or Nanobot
- Optional: Ollama running locally with nomic-embed-text model for hybrid vector+graph search
Install Mnemon
Install via Homebrew on macOS/Linux for the simplest setup. Alternatively install via Go or build from source.
# Homebrew (recommended):
brew install mnemon-dev/tap/mnemon
# Go install:
go install github.com/mnemon-dev/mnemon@latest
# Verify:
mnemon --versionRun the framework setup command
Use mnemon setup to auto-configure the MCP server in your agent framework. Mnemon will write the necessary config for the selected target.
# For Claude Code:
mnemon setup
# For other frameworks:
mnemon setup --target codex --yes
mnemon setup --target openclaw --yes
mnemon setup --target nanobot --global --yesConfigure data directory (optional)
By default Mnemon stores data in ~/.mnemon. Override with the MNEMON_DATA_DIR environment variable to use a custom path.
export MNEMON_DATA_DIR=~/projects/my-project/.mnemon
export MNEMON_STORE=my-project-storeEnable optional vector search with Ollama
For hybrid graph + vector recall, start Ollama with the nomic-embed-text model and configure Mnemon's embedding endpoint.
ollama pull nomic-embed-text
export MNEMON_EMBED_ENDPOINT=http://localhost:11434
export MNEMON_EMBED_MODEL=nomic-embed-textVerify Mnemon is active in your agent
Start Claude Code or your configured agent and ask it to remember a fact. Then in a new session, ask it to recall that fact to confirm cross-session persistence.
Mnemon Examples
Client configuration
Manual Claude Desktop config for Mnemon if not using the mnemon setup command.
{
"mcpServers": {
"mnemon": {
"command": "mnemon",
"args": ["serve"],
"env": {
"MNEMON_DATA_DIR": "/Users/you/.mnemon",
"MNEMON_STORE": "default"
}
}
}
}Prompts to try
Example prompts that exercise Mnemon's remember, link, and recall capabilities.
- "Remember that the auth service uses JWT tokens with a 24-hour expiry"
- "Link the database schema decision to the performance issue we discussed last week"
- "Recall everything you know about our API rate limiting approach"
- "What facts do you have stored about the payment service?"
- "Forget the outdated note about using Redis for session storage"Troubleshooting Mnemon
mnemon command not found after installation
For Homebrew installs, run: brew link mnemon. For go install, ensure $(go env GOPATH)/bin is in your PATH: export PATH=$PATH:$(go env GOPATH)/bin.
Facts remembered in one session are not recalled in the next
Verify MNEMON_DATA_DIR is consistent across sessions and is not a temp directory. Run mnemon setup again if the config may have pointed to a different store. Check the active store with: echo $MNEMON_STORE.
Vector search is not working or returns only graph results
Ensure Ollama is running (ollama serve) and the nomic-embed-text model is downloaded (ollama pull nomic-embed-text). Verify MNEMON_EMBED_ENDPOINT=http://localhost:11434 is set in the environment.
Frequently Asked Questions about Mnemon
What is Mnemon?
Mnemon is a Model Context Protocol (MCP) server that llm-supervised persistent memory for ai agents — graph-based recall, cross-session knowledge, single binary. works with claude code, openclaw, and any cli agent. It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Mnemon?
Follow the installation instructions on the Mnemon GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Mnemon?
Mnemon works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Mnemon free to use?
Yes, Mnemon is open source and available under the MIT license. You can use it freely in both personal and commercial projects.
Mnemon Alternatives — Similar Knowledge & Memory Servers
Looking for alternatives to Mnemon? Here are other popular knowledge & memory servers you can use with Claude, Cursor, and VS Code.
MemPalace
★ 52.6kA local AI memory system that stores all conversations verbatim and organizes them into navigable structures. It provides 19 MCP tools for AI assistants to search and retrieve past decisions, debugging sessions, and architecture debates automatically
Kratos
★ 25.7k🏛️ Memory System for AI Coding Tools - Never explain your codebase again. MCP server with perfect project isolation, 95.8% context accuracy, and the Four Pillars Framework.
Context Mode
★ 15.4kAn MCP server that preserves LLM context by intercepting large data outputs and returning only concise summaries or relevant sections. It enables efficient sandboxed code execution, file processing, and documentation indexing across multiple programm
Memu
★ 13.7kMemory for 24/7 proactive agents like OpenClaw.
MemOS
★ 9.3kMemOS (Memory Operating System) is a memory management operating system designed for AI applications. Its goal is: to enable your AI system to have long-term memory like a human, not only remembering what users have said but also actively invoking, u
Everos
★ 5.4kBuild, evaluate, and integrate long-term memory for self-evolving agents.
Browse More Knowledge & Memory MCP Servers
Explore all knowledge & memory servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Mnemon in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Mnemon?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.