Awesome MCP › Software Development

datalayer/jupyter-mcp-server

⭐ 1278 Python added to this list on 2025-04-13 repository created 2025-02-12

Jupyter MCP Server is an implementation of the Model Context Protocol (MCP) designed to enable real-time interaction and automation with Jupyter notebooks within JupyterLab or Notebook environments version 7 and above. It supports both local Jupyter instances and hosted notebooks via Datalayer, providing a seamless integration for users who want to control and interact with their notebooks programmatically. The server offers real-time control, allowing users to instantly view changes in notebooks as they happen, which is particularly useful for collaborative and dynamic data science workflows. It also features smart execution capabilities that automatically adjust when a cell run fails by leveraging cell output feedback, enhancing the robustness and reliability of notebook automation. The server is compatible with any MCP client, including popular ones like Claude Desktop AI assistant, Cursor, Windsurf, and others, making it versatile and adaptable to various user preferences and environments. The MCP server exposes multiple tools such as inserting and executing code cells, appending markdown cells, retrieving notebook information, and reading cell content, enabling advanced and fine-grained interactions with Jupyter notebooks. The project provides comprehensive documentation for setup, including environment preparation, starting JupyterLab with appropriate tokens and ports, and configuring MCP clients across different operating systems. It supports different transport mechanisms like Streamable HTTP and stdio, catering to diverse deployment scenarios. The project is actively maintained by Datalayer, Inc., and includes resources such as blog posts and videos to help users get the most out of the server. Overall, Jupyter MCP Server facilitates enhanced automation, control, and integration of Jupyter notebooks through the Model Context Protocol, making it a valuable tool for developers and data scientists working with interactive computational notebooks.

https://github.com/datalayer/jupyter-mcp-server

aiai-assistantappend_markdown_cellautomationcell-output-feedbackclaude-desktopcode-executioncollaborative-workflowscursordata-science-workflowsdatalayerdockerget_notebook_infoinsert_execute_code_cellintegrationjupyterjupyter-notebooksjupyterlabmarkdown-cellsmcpmcp-clientsmcp-servermodel-context-protocolnotebook-automationnotebook-controlread_cellreal-time-collaborationreal-time-interactionserversmart-executionsmitherystdio-transportstreamable-httptoolswindsurf