Openml MCP Server

🚀 OpenML MCP Server

An MCP (Model Context Protocol) server that allows clients (such as Claude Desktop) to interact with the public OpenML API. This server exposes various endpoints of the OpenML API as MCP tools, enabling users to query datasets, tasks, flows, runs, and more directly from MCP-compatible clients.

🚀 Quick Start

For End-Users

These instructions describe how to use this server with an MCP client (e.g., Claude Desktop).

Prerequisites:

1. Claude Desktop: Ensure you have the latest version installed (Download Link).
2. uv: Used to run the server. Install it via the following command:

pip install uv

For Developers

To run the server:

1. Install the dependencies:

pip install openml-mcp-server[mongo]

2. Configure the environment variables:

export OPENML_MCP_SERVER__MONGO_URI=mongodb://localhost:27017/

3. Start the server:

uv openml-mcp-server

✨ Features

• Provide MCP tools corresponding to OpenML GET endpoints.
• Query datasets, tasks, flows, runs, and evaluations (e.g., support API filtering).
• Built using the mcp Python SDK (based on FastMCP).

📦 Installation

End-Users

Prerequisites:

• Claude Desktop: Make sure the latest version is installed (Download).
• uv: Install it using the command pip install uv.

Developers

1. Install dependencies:

pip install openml-mcp-server[mongo]

2. Set environment variables:

export OPENML_MCP_SERVER__MONGO_URI=mongodb://localhost:27017/

3. Start the server:

uv openml-mcp-server

💻 Usage Examples

Basic Usage

For End-Users

1. Query Datasets:

openml-mcp-server dataset --name " mnist"

2. Create a New Task:

openml-mcp-server task create --name "my_task" --description "A test task."

Advanced Usage

For developers, after starting the server, you can perform more complex operations by customizing environment variables and configurations.

📚 Documentation

Configuration

To modify the configuration file of Claude Desktop:

1. Open ~/.config/claude/settings.json.
2. Add the following content:

{
  "openml-mcp-server": {
    "api-key": "your_api_key_here"
  }
}

3. Set environment variables to control server behavior:

export OPENML_MCP_SERVER__DEBUG=true

Development

To contribute code:

1. Clone the repository:

git clone https://github.com/yourusername/openml-mcp-server.git

2. Create a feature branch and commit your changes.
3. Submit a pull request.

Running Tests

To run tests on the server:

pytest tests/

Contribution

To contribute, please refer to CONTRIBUTING.md.

📄 License

This project is open-source under the MIT License. For details, see LICENSE.