memoh-mcp

MCP server for managing Memoh AI agents from Claude Code, Cursor, or any MCP-compatible client.

26 Tools

Related MCP server: Spaceship MCP

Quick Start

1. Install

pip install mcp python-dotenv

Or with uv:

uv pip install mcp python-dotenv

2. Configure

Set environment variables (or create a .env file next to the server script):

export MEMOH_API=http://localhost:8080
export MEMOH_USER=admin
export MEMOH_PASS=your-password

3. Add to Claude Code

Add to ~/.claude/settings.json:

{
  "mcpServers": {
    "memoh": {
      "command": "python3",
      "args": ["/path/to/memoh_mcp_server.py"]
    }
  }
}

Or with a .env wrapper script:

{
  "mcpServers": {
    "memoh": {
      "command": "/path/to/wrapper.sh"
    }
  }
}

Where wrapper.sh:

#!/bin/bash
exec python3 /path/to/memoh_mcp_server.py "$@"

4. Add to Cursor

Add to .cursor/mcp.json:

{
  "mcpServers": {
    "memoh": {
      "command": "python3",
      "args": ["/path/to/memoh_mcp_server.py"],
      "env": {
        "MEMOH_API": "http://localhost:8080",
        "MEMOH_USER": "admin",
        "MEMOH_PASS": "your-password"
      }
    }
  }
}

Usage

Once connected, your AI assistant can manage Memoh bots directly:

• List bots: See all bots with their IDs and status

• Read/write files: Edit bot identity, soul, memory, and other config files in containers

• Manage skills: Add procedural protocols that guide bot behavior

• Vector memory: Add, search, and delete bot memories

• Chat: Send messages to bots and read their responses

• Settings: Configure models, heartbeat intervals, and other bot settings

Requirements

• Python 3.10+

• Running Memoh server

• mcp Python package (MCP SDK)

• python-dotenv (optional, for .env file support)

License

MIT