jupyterlab_nb_venv_kernels_ui_extension

Context menu extension for kernel launcher cards. Right-click on any kernel to navigate to its project directory, open a terminal, or manage the virtual environment. Intended to complement nb_venv_kernels for workspace management.

Features

Context Menu (right-click on kernel launcher card):

• Show in File Browser - Navigate to the kernel's project root
• Open Terminal at Location - Open terminal at the kernel's project directory
• Unregister Kernel - Remove kernel from registry (requires nb_venv_kernels)
• Remove Environment (dangerous) - Permanently delete local .venv environments with confirmation (requires nb_venv_kernels)

Kernel Menu (provided by nb_venv_kernels):

• Scan for Virtual Environments - Discover and register new virtual environments in your workspace

General:

• Project-aware navigation - For .venv environments, navigates to project root (one level up from .venv)
• Dynamic kernel support - Works with nb_conda_kernels and nb_venv_kernels providers
• Kernel details on hover - The card's tooltip shows kernel name, kind, executable, resource directory and environment path
• Shared display names refused - When kernels in different directories share one display name, every menu action refuses and names the colliding kernels

Requirements

• JupyterLab >= 4.0.0

Install

pip install jupyterlab-nb-venv-kernels-ui-extension

Agent Skill

.agents/skills/jupyterlab-nb-venv-kernels-ui-extension/SKILL.md tells an AI coding assistant how to manage these kernels from the shell with the nb_venv_kernels CLI. pip install puts a copy in share/jupyter/agents/skills/ under the Python environment. No agent reads that directory, so link it into the agent skills directory. Run this where python is the Python that runs the lab, not an activated project venv:

d="$(python -c 'import sysconfig; print(sysconfig.get_paths()["data"])')/share/jupyter/agents/skills/jupyterlab-nb-venv-kernels-ui-extension"
ls "$d/SKILL.md" >/dev/null &&
  mkdir -p ~/.agents/skills &&
  ln -sfn "$d" ~/.agents/skills/jupyterlab-nb-venv-kernels-ui-extension

When there is no copy under that Python's data directory, ls prints the missing path, the command stops with a non-zero exit, and any existing link stays. That happens in a project venv, and after pip install --user; for that install, put python -m site --user-base in place of the python -c '...' call in the first line. Claude Code reads ~/.claude/skills, not ~/.agents/skills; for Claude Code, put ~/.claude/skills in place of ~/.agents/skills in the last two lines.

Agents that read .agents/skills also find the skill in a clone of this repository. To make it available to Claude Code everywhere, link it from the root of a clone:

mkdir -p ~/.claude/skills &&
  ln -sfn "$PWD/.agents/skills/jupyterlab-nb-venv-kernels-ui-extension" ~/.claude/skills/jupyterlab-nb-venv-kernels-ui-extension

When a copied directory, not a link, already sits at the link path of either command, ln puts the new link inside it and the agent keeps reading the copy; move that directory out of the skills directory, then run the command again.

Uninstall

pip uninstall jupyterlab_nb_venv_kernels_ui_extension