Advertise
MCP.so
Sign In

Jupyter MCP Server for Claude Desktop

@harshitha-8

Repository

Basic information

Category

Data & Analytics

Transports

stdio

Author

harshitha-8

About Jupyter MCP Server for Claude Desktop

Jupyter MCP Server Jupyter MCP Server is an implementation of the Model Context Protocol (MCP) server that enables interaction with Jupyter notebooks running in any JupyterLab environment, including your local JupyterLab instance

#jupyter#mcp#claude-desktop

Jupyter MCP Server for Claude Desktop config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "Integrating-the-Jupyter-server-with-claude-desktop-uisng-the-powerful-model-context-protocol": {
      "command": "npx",
      "args": [
        "-y",
        "@smithery/cli",
        "install",
        "@datalayer/jupyter-mcp-server",
        "--client",
        "claude"
      ]
    }
  }
}

Jupyter MCP Server for Claude Desktop tools

Tool list not available yet

This listing does not currently have a tool list to display. This does not mean the server has no tools; check its documentation for supported capabilities.

Fetch the live tool list by running this server in a temporary sandbox using the button above.

Overview

What is Jupyter MCP Server for Claude Desktop?

Jupyter MCP Server for Claude Desktop is an implementation of the Model Context Protocol (MCP) server that enables interaction with Jupyter notebooks running in any JupyterLab environment, including local instances. It is designed for seamless integration with Claude Desktop across macOS, Windows, and Linux.

How to use Jupyter MCP Server for Claude Desktop?

Install the server Docker image or use Smithery, configure it in Claude Desktop’s MCP settings with environment variables SERVER_URL, TOKEN, and NOTEBOOK_PATH, then start JupyterLab on port 8888. The server provides two tools: add_execute_code_cell and add_markdown_cell.

Key features of Jupyter MCP Server for Claude Desktop

  • Provides an MCP server interface for Jupyter notebooks.
  • Supports real-time collaboration via JupyterLab.
  • Integrates smoothly with Claude Desktop on all platforms.
  • Programmatically add and execute code or markdown cells.
  • Runs inside Docker for easy deployment.

Use cases of Jupyter MCP Server for Claude Desktop

  • Let Claude interactively execute Python code in a Jupyter notebook.
  • Automate notebook cell creation and execution from AI conversations.
  • Enable collaborative data analysis with AI assistance.
  • Integrate JupyterLab workflows with Claude Desktop’s MCP capabilities.

Frequently asked questions

1

What dependencies are required to run the server?

You need JupyterLab 4.4.1, jupyter-collaboration 4.0.2, ipykernel, and datalayer_pycrdt 0.12.15. The server itself runs in Docker.

2

How do I configure the server for Linux vs macOS/Windows?

On Linux use `--network=host` and `http://localhost:8888`; on macOS/Windows use `http://host.docker.internal:8888`. The `TOKEN` must match your JupyterLab setup.

3

Can I install the server without Docker?

The README only describes Docker usage and Smithery installation via `npx`. No native install is documented.

4

What tools does the server expose?

Two tools: `add_execute_code_cell` (takes `cell_content` string, returns cell output) and `add_markdown_cell` (takes `cell_content` string, returns success message).

5

Where does the notebook file live?

The `NOTEBOOK_PATH` environment variable defines the notebook file name (e.g., `notebook.ipynb`), relative to the directory where JupyterLab is started.

Comments

More Data & Analytics MCP servers

Lawstronaut Legal Research

Lawstronaut-FZCO

Connect your AI to millions of laws and court cases! Lawstronaut is the infrastructure layer that connects AI agents and software to millions of legal documents across 150+ jurisdictions, structured legal data, continuou

1

Grabbit MCP

Grabbit

Remote MCP server that helps AI agents find Reddit buyer-intent threads, check subreddit rules, read full conversations, and manage human-reviewed lead workflows.

World Monitor

Elie Habib

Live global-intelligence MCP server covering news, markets, conflicts, aviation, maritime activity, energy, climate, cyber threats, supply chains, and source-grounded regional briefs.

OrbitWan.io

Wanchain explorer data and analytics: addresses, transactions, tokens, validators, cross chain transfers, plus OrbitPay prepaid premium data (x402).

Forency

cheevahagadog

Investigate any website's technology stack: CMS, eCommerce, frameworks, analytics, CRM, payments, and hosting behind any URL. Ask "which of these 15 prospects use HubSpot vs Salesforce?" and get a table back. Free tier:

1
Advertise