{"categories":[{"label":"Application Frameworks","url":"https://skillfed.io/packages/category/software-development-libraries-application-frameworks/5"}],"enrichment":{"capability":"Exposes Python functions as tools to MCP clients via a Jupyter Server extension, enabling AI assistants and other MCP-compatible applications to call registered functions through an HTTP endpoint.","skillfed_tags":["jupyter-extension","mcp-protocol","ai-integration"],"use_cases":["Enable Claude or other MCP clients to read and edit Jupyter notebooks by registering notebook toolkit functions as MCP tools","Expose custom Python utility functions to AI assistants running in a terminal or IDE without modifying the assistant's codebase","Build a bridge between Jupyter-based data analysis and AI coding agents that need access to notebook operations and file system tools","Register standard library functions (os, json, time) or third-party tools as MCP endpoints for agent automation workflows","Allow package authors to expose their tools to MCP clients by declaring entrypoints in pyproject.toml"],"what_it_does":"jupyter-server-mcp is a Jupyter Server extension that bridges Jupyter and the Model Context Protocol (MCP), allowing you to register Python functions as tools accessible to MCP clients like Claude, Mistral, and other AI assistants. It runs an HTTP server (default port 3001) that exposes registered functions through the MCP protocol, making Jupyter functionality available to external AI tools.\n\nThe extension supports two registration patterns: manual configuration via Jupyter's traitlets system (specifying tools as \"module:function\" strings) and automatic discovery through Python package entrypoints. It depends on fastmcp for the underlying MCP server implementation and integrates directly with Jupyter Server's extension lifecycle, starting the MCP server automatically when Jupyter starts.","worth_installing":"Yes, if you run Jupyter Server and want to make its functionality available to MCP clients. The extension is actively maintained, has low install friction, carries a permissive license, and requires only two common dependencies. Start with the quick-start configuration to expose a few functions and verify the MCP endpoint works with your chosen client."},"id":"jupyter-server-mcp","links":{"html":"https://skillfed.io/packages/jupyter-server-mcp","md":"https://skillfed.io/packages/jupyter-server-mcp.md","pypi":"https://pypi.org/project/jupyter-server-mcp/"},"maintenance":{"status":"active"},"meta":{"latest_release":"2026-04-23","license_spdx":null,"license_treatment":"permissive","name":"jupyter-server-mcp","python_support":"supports_current","summary":null},"popularity":{"monthly_downloads":84804,"position":13972,"tier":"top_15000"},"security":{"n_vulnerabilities":0},"version":"0.2.1"}