MindManager MCP Server

A Model Context Protocol (MCP) server implementation for the mindm library, providing a standardized interface to interact with MindManager on Windows and macOS.

PyPI version

Overview

This server allows you to programmatically interact with MindManager through the Model Context Protocol (MCP), a standardized way to provide context and tools to LLMs. It leverages the mindm library to manipulate MindManager documents, topics, relationships, and other mindmap elements.

Example:

MindManager MCP in Claude

Features

Retrieve mindmap structure and central topics
Export mindmaps to Mermaid, Markdown, JSON formats to be used in LLM chats
Get information about MindManager installation and library folders
Get current selection from MindManager

Planned Features

Create new mindmaps from serialized data
Add, modify, and manipulate topics and subtopics
Add relationships between topics
Add tags to topics
Set document background images

Requirements

Python 3.12 or higher
mcp package (Model Context Protocol SDK)
mindm library (included in this project)
MindManager (supported versions: 23-) installed on Windows or macOS

Installation macOS

# Clone the repository (if you're using it from a repository)
git clone https://github.com/robertZaufall/mindm-mcp.git
cd mindm-mcp

# create a virtual environment for Python
brew install uv # if needed
uv pip install -r pyproject.toml

# alternative: manual installation of modules
uv add "mcp[cli]"
uv add fastmcp
uv add markdown-it-py
uv add -U --index-url=https://test.pypi.org/simple/ --extra-index-url=https://pypi.org/simple/ mindm mindm-mcp

Installation Windows

# Change to DOS command prompt
cmd

# Clone the repository (if you're using it from a repository)
git clone https://github.com/robertZaufall/mindm-mcp.git
cd mindm-mcp

# create a virtual environment for Python
pip install uv # if needed
uv pip install -r pyproject.toml

# install nodejs
choco install nodejs # if you have chocolatey installed. If not install nodejs otherwise
refreshenv
node -v
npm install -g npx

Usage

MCP inspector

# run mcp with inspector
uv run --with mind --with fastmcp --with markdown-it-py mcp dev mindm_mcp/server.py

Claude Desktop

Local python file

Adjust the path for the local file as needed.

{
  "mcpServers": {
    "mindm (MindManager)": {
      "command": "uv",
      "args": [
        "run",
        "--with",
        "mindm>=0.0.4.6",
        "--with",
        "fastmcp",
        "--with",
        "markdown-it-py",
        "/Users/master/git/mindm-mcp/mindm_mcp/server.py"
      ]
    }
  }
}

Module from package repository

Adjust VIRTUAL_ENV as needed.

{
    "mcpServers": {
      "mindm (MindManager)": {
        "command": "uv",
        "args": [
          "run",
          "--with",
          "mindm>=0.0.4.6",
          "--with",
          "mindm-mcp>=0.0.1.50",
          "--with",
          "fastmcp",
          "--with",
          "markdown-it-py",
          "-m",
          "mindm_mcp.server"
        ],
        "env": {
            "VIRTUAL_ENV": "/Users/master/git/mindm-mcp/.venv"
        }
      }
    }
}

Hint: If the MCP server does not show up with the hammer icon on Windows, close Claude Desktop and kill all background processes.

MCP Tools

The server exposes the following tools through the Model Context Protocol:

Document Interaction

get_mindmap: Retrieves the current mindmap structure from MindManager
get_selection: Retrieves the currently selected topics in MindManager
get_library_folder: Gets the path to the MindManager library folder
get_grounding_information: Extracts grounding information (central topic, selected subtopics) from the mindmap

Serialization

serialize_current_mindmap_to_mermaid: Serializes the currently loaded mindmap to Mermaid format
serialize_current_mindmap_to_markdown: Serializes the currently loaded mindmap to Markdown format
serialize_current_mindmap_to_json: Serializes the currently loaded mindmap to a detailed JSON object with ID mapping

Platform Support

Windows: Full support for topics, notes, icons, images, tags, links, relationships, and RTF formatting
macOS: Support for topics, notes, and relationships (limited support compared to Windows)

Integration with Claude and other LLMs

This MCP server can be installed in Claude Desktop or other MCP-compatible applications, allowing LLMs to:

Access mindmap content
Manipulate mindmaps (coming)
Create new mindmaps based on LLM-generated content (coming)

Troubleshooting

Ensure MindManager is running before starting the server
For macOS, make sure you allow Claude Desktop to automate MindManager

Acknowledgements

This project is built upon the mindm library, providing Python interfaces to MindManager on Windows and macOS platforms. It uses the Model Context Protocol (MCP) SDK developed by Anthropic.

License

MIT License - See LICENSE file for details