JupyterMCP
by jjsantos01
JupyterMCP connects Jupyter Notebook to Claude AI through the Model Context Protocol (MCP), allowing Claude to directly interact with and control Jupyter Notebooks. This integration enables AI-assisted code execution, data analysis, visualization, and more.
Last updated: N/A
What is JupyterMCP?
JupyterMCP is a tool that integrates Jupyter Notebook version 6.x with Claude AI using the Model Context Protocol (MCP). It establishes a two-way communication channel, allowing Claude to interact with and control the Jupyter Notebook environment.
How to use JupyterMCP?
To use JupyterMCP, you need to install the required prerequisites, including Python 3.12+, the uv package manager, and the Claude AI desktop application. After cloning the repository and creating a virtual environment, you need to configure Claude to use the JupyterMCP server. Then, start the Jupyter Notebook server with the jupyter-mcp kernel and initialize the WebSocket server within the notebook.
Key features of JupyterMCP
Two-way communication between Claude AI and Jupyter Notebook
Cell manipulation (insert, execute, manage)
Notebook management (save, retrieve info)
Cell execution (specific or all cells)
Output retrieval (text and images)
Use cases of JupyterMCP
AI-assisted code generation and execution
Automated data analysis workflows
Interactive data visualization with AI assistance
Creating presentations with AI-driven content
Running Stata code within Jupyter Notebook via Claude
FAQ from JupyterMCP
What Jupyter Notebook versions are supported?
What Jupyter Notebook versions are supported?
Only Jupyter Notebook version 6.x is supported. Jupyter Lab, Jupyter Notebook v7.x, VS Code Notebooks, and Google Colab are not compatible.
What are the prerequisites for installation?
What are the prerequisites for installation?
You need Python 3.12 or newer, the uv package manager, and the Claude AI desktop application.
How do I configure Claude to use JupyterMCP?
How do I configure Claude to use JupyterMCP?
You need to edit the claude_desktop_config.json file in Claude's settings to include the JupyterMCP server configuration with the correct path to the src folder.
What tools are available to Claude once connected?
What tools are available to Claude once connected?
Claude has access to tools like ping, insert_and_execute_cell, save_notebook, get_cells_info, get_notebook_info, run_cell, run_all_cells, get_cell_text_output, get_image_output, edit_cell_content, and set_slideshow_type.
What are the limitations of JupyterMCP?
What are the limitations of JupyterMCP?
It only supports Jupyter Notebook 6.x, text output from cells is limited to 1500 characters, it doesn't support advanced Jupyter widget interactions, and the connection may timeout after periods of inactivity.