jupyter-live-kernel

A stateful live Jupyter kernel keeps Python variables persistent across multiple code runs by maintaining in-memory context. This allows variables, imports, and objects to persist across iterative execution steps without reloading.

Executing Python code repeatedly against a live kernel is the best way to handle iterative exploratory data science. It keeps the same in-memory context, enabling step-by-step feature engineering and intermediate variable inspection.

Yes, a running JupyterLab instance is required to use a stateful Python REPL. You must create a scratch notebook for a kernel session before executing Python code repeatedly against it.

Yes, you can edit Jupyter notebook cells on the fly during an interactive Python session. The live kernel integration supports adding, replacing, or deleting cells to evolve your workflow incrementally.

You inspect variables and enumerate running notebooks in Jupyter through the live kernel's session discovery commands. These commands list running servers, enumerate notebooks, and provide structured JSON outputs for variable previews.