Jupyter Server Provider

The Jupyter Server provider is the default. Code executes through the code-sandboxes jupyter variant against a local or remote Jupyter Server (SANDBOX_VARIANT=jupyter), giving you a full Jupyter kernel with process isolation, persistent state, and rich outputs (images, plots, HTML).

tip

This engine uses the code-sandboxes Jupyter engine. For other SANDBOX_VARIANT values (kaggle, colab, monty, modal, ...), see the code-sandboxes documentation for engine-level parameters and credential setup. Non-jupyter engines also require the optional jupyter_mcp_sandboxes extension (pip install jupyter_mcp_sandboxes).

Requirements

pip install jupyter-mcp-server

Start a Jupyter Server (or JupyterLab):

jupyter lab --port 8888 --IdentityProvider.token MY_TOKEN --ip 0.0.0.0

Configuration

Variable	Description	Default
`JUPYTER_URL`	Jupyter Server URL (used for both document and code sandbox)	`http://localhost:8888`
`JUPYTER_TOKEN`	Jupyter Server authentication token	`None`
`DOCUMENT_ID`	Default notebook path (optional)	`None`

MCP client configuration:

{
  "mcpServers": {
    "jupyter": {
      "command": "uvx",
      "args": ["jupyter-mcp-server@latest"],
      "env": {
        "JUPYTER_URL": "http://localhost:8888",
        "JUPYTER_TOKEN": "MY_TOKEN"
      }
    }
  }
}

For advanced deployments that separate notebook storage from kernel execution, use DOCUMENT_URL / DOCUMENT_TOKEN and CODE_SANDBOX_URL / CODE_SANDBOX_TOKEN individually. See the Configuration reference for the full list of variables.

Requirements​

Configuration​

Requirements

Configuration