notebook-refresh

name	notebook-refresh
description	Clears and re-runs Jupyter notebooks using the project's uv environment. Use when the user wants to validate, refresh, or "clean and run" notebooks.

Notebook Refresh Skill

Instructions

1. Determine Scope: Decide which notebooks to process based on the user's request:
   • All notebooks: Find every .ipynb file in the specified directory.
   • Changed notebooks only: If the user asks to refresh only changed/modified notebooks, use git to identify them:
     • git diff --name-only HEAD -- '*.ipynb' for uncommitted changes vs the last commit.
     • git diff --name-only main...HEAD -- '*.ipynb' for all changes on the current branch vs main (adjust base branch name if needed).
     • If no notebooks have changed, report that and stop.
2. Prerequisites: Ensure the uv environment has nbconvert and ipykernel installed. If not, prompt to run uv add --dev nbconvert ipykernel.
3. Execution: For each notebook, run the following commands from the project root:
   • Step 1 (Clear): uv run jupyter nbconvert --ClearOutputPreprocessor.enabled=True --inplace <path_to_notebook>.ipynb
   • Step 2 (Run): uv run jupyter nbconvert --to notebook --execute --inplace --ExecutePreprocessor.timeout=600 <path_to_notebook>.ipynb
4. Validation:
   • If a notebook fails, capture the error message from the terminal output.
   • Report exactly which cell or module caused the failure.
5. Final Report: Provide a summary table of results (Success/Failure) for all processed files, noting whether the scope was "all" or "changed only".