JupyterLab
chirashi.jupyterlab
A Docker Compose file for JupyterLab alone, generated from the catalog and checked by the safety inspector. No account needed.
Browser-based development environment presenting notebooks, code editors, terminals and data files in one tabbed workspace. This image variant ships a Python data-science stack preinstalled.
JupyterLab arranges notebooks, text editors, terminals, file browsers and output views as panels that can be placed side by side, with a kernel process per notebook executing code and returning results inline. Kernels exist for many languages, and the extension system adds debuggers, variable inspectors, language servers and version control integration. The data-science image variant arrives with the common numerical, statistical and plotting libraries already installed, which removes the first hour of environment assembly. A first deployment mounts a work directory, opens the interface using the token printed at startup and runs a notebook cell.
You know it worked when
- The interface loads using the token from the container log.
- A notebook cell runs and returns output.
- A preinstalled library imports without an installation step.
- A file saved into the mounted directory is present on the host.
- Work in that directory survives a container restart.
Known sharp edges
- Only the mounted work directory persists; packages installed at runtime and files written elsewhere disappear when the container is recreated.
- Access is controlled by a token printed to the container log, and disabling it for convenience leaves arbitrary code execution open to anyone reaching the port.
- A notebook has full access to the container, so it is a shell by another name and should be treated with the same caution.
- The Docker Hub jupyter namespace has been superseded by a different registry and is no longer the actively updated source for these images.