jupyter-live-kernel

Use a live Jupyter kernel as a stateful REPL. Start a JupyterLab server, create a kernel session via the Jupyter REST API, then execute code against the same notebook so variables, imports, and objects persist across calls.

Use a live kernel for iterative exploration, data science, and tasks where you build state incrementally. Use one-shot execution for standalone scripts that need no persisted variables, and a terminal for shell commands, installs, and git operations.

Run the variables subcommand with the notebook path to list all live variables, or pass a variable name to preview its value. This works because the kernel retains its full namespace between executions.

The kernel needs time to initialize after the server starts, so the first execution may exceed the default 30-second timeout. Retry the execution or pass a longer timeout such as 60 or 120 seconds for setup and heavy computation.

The kernel uses JupyterLab's Python environment, so packages must be installed into that environment before use. Install them into the JupyterLab tool environment first, then import them in your executed code.