uvk 0.0.1

Fully isolated IPython kernels based on uv

uvk -- Fully isolated IPython kernels based on uv

Proper documentation TBD... In the meantime:

1. Install uvk in an environment from which you run Jupyter Lab / Notebook.
2. Invoke uvk from this environment to install the UVK kernel attached to the Python executable of your environment.
   • Distinct icon TBD
3. Use kernel UVK (Python 3.xx) in your notebooks

and 🎉 your Python kernel runs out of a temporary, isolated environment. Check it out by running a notebook with a code cell going

import sys
sys.prefix

and see how every time you restart the kernel, the prefix directory changes.

Feature wishlist:

• Python version validation against a constraint with a magic
• Magic for adding dependencies on the fly
• Magic for including inline script metadata
• Integration with uv project workflows
• Python version and executable management through duly-named alternate kernels