mcp-ploomber

Setup:

# requires conda
pip install invoke
invoke setup
conda activate mcp-ploomber

If you're not using conda, create a virtual env and do:

pip install --editable .

Virtual Environment Setup

You'll need to set up a virtual environment to run the MCP server. We provide setup scripts to make this easy:

1. Run the setup script to create and configure the virtual environment:

# Make the script executable first
chmod +x setup_venv.sh
./setup_venv.sh

This script will:

• Create a Python virtual environment
• Install all required dependencies
• Set up the package in development mode

MCP Configuration

To configure the MCP server, you need to create an mcp.json file:

1. Run the generation script to create a personalized configuration:

# Make the script executable first
chmod +x generate_mcp_json.sh
./generate_mcp_json.sh

2. Edit the generated mcp.json file to add your Ploomber Cloud API key:

{
    "mcpServers": {
        "ploomber-mcp": {
            "command": "/path/to/your/venv/bin/python",
            "args": [
                "/path/to/your/src/mcp_ploomber/server.py"
            ],
            "env": {
                "_PLOOMBER_CLOUD_ENV": "",
                "PLOOMBER_CLOUD_KEY": "YOUR_API_KEY_HERE"
            },
            "setup": "/path/to/your/setup_venv.sh"
        }
    }
}

Running the Server

Once you've completed the setup, you can start the MCP server manually using:

cd src/mcp_ploomber

mcp run server.py

To debug, you can also run the MCP server in dev mode:

mcp dev server.py

To add the MCP to your LLM client, simply copy/paste your mcp.json where the client expects the MCP config file.