cursor-notebook

The Jupyter Notebook MCP Server (cursor-notebook) exposes a rich set of MCP tools that let AI agents read, modify, and manage Jupyter notebooks (.ipynb) directly through a standardized API. Once running, you can create, read, edit, and export notebooks, manipulate individual cells, work with cell metadata and outputs, and even bulk-add cells. The server is designed to safely operate within user-defined directories, leveraging nbformat for notebook structure handling and nbconvert for exporting notebooks to formats like Python scripts and HTML. Tools such as notebook_read, notebook_edit_cell, notebook_add_cell, notebook_export, notebook_get_outline, and notebook_search enable a practical and scriptable workflow for notebook automation from an agent such as Cursor or other MCP-enabled models.

Prerequisites:

• Python 3.10+ installed on your system
• pip available in your environment

Step-by-step installation:

1. Create and activate a Python virtual environment (optional but recommended): python3 -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate

2. Install the package from PyPI: pip install cursor-notebook-mcp

3. Run the MCP server (example using uvx with the package name as the entry point): uvx cursor-notebook-mcp

4. Verify the server is running (default port and transport depend on configuration). If you configured environment variables, ensure they’re set before starting the server.

5. (Optional) Create a dedicated configuration file or environment to set allowed roots, export formats, and transport options as described in the README.

Notes and tips:

• The server restricts operations to user-defined directories for safety. Configure ALLOWED_ROOTS to limit file access.
• nbformat is used for notebook manipulation; nbconvert enables exporting to formats such as Python scripts and HTML. Ensure external dependencies (Pandoc/LaTeX) are installed if you plan to export to PDF or other formats.
• If you enable SFTP support or Streamable HTTP transport, you may need additional system dependencies and network configurations.
• When upgrading, check for compatibility with pydantic versions (noted in changelog 0.3.2+ mentions pydantic 2.12.0+ compatibility adjustments).
• The server exposes a broad suite of tools; refer to notebook_mcp tool list in the README for exact method names and usage patterns.