Jupyter MCP Server — MCP reference

Jupyter MCP Server is a Model Context Protocol server that lets AI agents operate Jupyter notebooks: managing notebooks and cells, executing code on live kernels, and provisioning code sandboxes.

This reference documents the server's complete MCP surface — 22 tools and 1 prompt (protocol revision 2025-11-25) — captured from a running server built from this repository. Each page shows the exact schema the server advertises plus a link to the decorator that registers it. The MCP Reference tab renders the same snapshot as a single interactive page, and Configuration covers every runtime setting and both transports.

Connection & server

Tool	Summary
`connect_to_jupyter`	Connect to a Jupyter server dynamically with URL and token.
`list_files`	List all files and directories recursively in the Jupyter server's file system.
`list_kernels`	List all available kernels in the Jupyter server.

Notebooks

Tool	Summary
`use_notebook`	Use a notebook and activate it for following cell operations.
`list_notebooks`	List all notebooks that have been used via use_notebook tool
`read_notebook`	Read a notebook and return index, source content, type, execution count of each cell.
`restart_notebook`	Restart the kernel for a specific notebook.
`unuse_notebook`	Unuse from a specific notebook and release its resources.

Cells

Tool	Summary
`insert_cell`	Insert a cell to specified position from the currently activated notebook.
`read_cell`	Read a specific cell from the currently activated notebook and return it's metadata (index, type, execution count), source and outputs (for code cells)
`edit_cell_source`	Perform a surgical find-and-replace within a cell's source (like an editor's Edit tool).
`overwrite_cell_source`	Replace the entire source of a cell in the currently activated notebook.
`move_cell`	Move a cell from source_index to target_index within the currently activated notebook.
`delete_cell`	Delete specific cells from the currently activated notebook and return the cell source of deleted cells (if include_source=True).
`clear_cell_output`	Clear the outputs and execution count of a single code cell in the currently

Execution

Tool	Summary
`execute_cell`	Execute a cell from the currently activated notebook with timeout and return it's outputs
`insert_execute_code_cell`	Insert a cell at specified index from the currently activated notebook and then execute it with timeout and return it's outputs
`execute_code`	Execute code directly in a kernel (not saved to notebook).

Sandboxes (extension)

Tool	Summary
`launch_sandbox`	Launch a code sandbox that can be used instead of Jupyter kernels.
`list_sandboxes`	List launched code sandboxes that can be used as alternatives to kernels.
`use_sandbox`	Select which launched sandbox execute_code should use instead of kernels.
`terminate_sandbox`	Terminate a launched code sandbox.

Prompts

Prompt	Summary
`jupyter_cite`	Like @ or # in Coding IDE or CLI, cite specific cells from specified notebook and insert them into the prompt.

About these docs

Generated with Sourcey from an mcp-parser stdio snapshot of the server (jupyter-mcp-server --transport stdio --start-new-code-sandbox false). The snapshot, source map, and page generator are checked in under docs/sourcey/, and CI regenerates them on every pull request and fails on any difference, so this page cannot drift from the code.