Datalayer Code Sandbox

The Jupyter MCP Server can execute code on the Datalayer cloud code sandbox, providing fully isolated execution with GPU support, snapshots, and persistence. Execution is routed through the code-sandboxes datalayer engine, which is powered by agent_code_sandboxes.

tip

For the full, engine-level credential and parameter reference, see the Datalayer sandbox guide in the code-sandboxes documentation.

Requirements

Install the Datalayer extra:

pip install "jupyter-mcp-server[datalayer]"

Install the sandbox extension package used by this sandbox:

pip install datalayer_mcp_sandboxes

Credentials

1. Create an account at datalayer.ai.
2. Generate an API token from your account settings (IAM → Tokens / API Keys).
3. Export it as an environment variable:

Environment variable	Description
DATALAYER_API_KEY	API key for Datalayer code sandbox authentication
DATALAYER_RUN_URL	Custom Datalayer service URL (optional, self-hosted)

Configuration

Select the Datalayer engine with SANDBOX_VARIANT=datalayer:

SANDBOX_VARIANT=datalayer
DATALAYER_API_KEY=your-datalayer-token

The engine reads DATALAYER_API_KEY itself. The server also accepts the sandbox-neutral spelling, which is what --code-sandbox-url and --code-sandbox-token set and what the other sandboxes use:

Environment variable	Description
CODE_SANDBOX_URL	Datalayer service URL, e.g. https://prod1.datalayer.run
CODE_SANDBOX_TOKEN	Datalayer API token
SANDBOX_ENVIRONMENT	Environment to launch, e.g. python-cpu-env

"env": {
  "SANDBOX_VARIANT": "datalayer",
  "CODE_SANDBOX_URL": "https://prod1.datalayer.run",
  "CODE_SANDBOX_TOKEN": "your-datalayer-token",
  "SANDBOX_ENVIRONMENT": "python-cpu-env"
}

MCP client configuration:

{
  "mcpServers": {
    "jupyter": {
      "command": "uvx",
      "args": ["jupyter-mcp-server@latest"],
      "env": {
        "SANDBOX_VARIANT": "datalayer",
        "DATALAYER_API_KEY": "your-datalayer-token"
      }
    }
  }
}