#What it does

The jupyter-notebook skill creates clean, reproducible Jupyter notebooks for two primary modes: experiments/exploratory analysis and tutorials/teaching walkthroughs. It uses bundled templates and a helper script to generate well-structured notebooks and avoid hand-authoring raw notebook JSON.

#How to use

$jupyter-notebook

Create a Jupyter notebook for this task with clear sections, runnable cells, and concise takeaways.

#Decision tree

• Exploratory, analytical, or hypothesis-driven -- choose experiment
• Instructional, step-by-step, or audience-specific -- choose tutorial
• Editing an existing notebook -- treat as a refactor

#Workflow

1. Lock the intent -- Identify kind (experiment or tutorial), objective, audience, and done criteria
2. Scaffold from template -- Use the helper script to avoid raw JSON authoring
3. Fill with small, runnable steps -- One step per code cell, markdown cells for explanations
4. Apply the right pattern -- Experiment patterns or tutorial patterns from references
5. Edit safely -- Preserve structure, prefer targeted edits over full rewrites
6. Validate -- Run top-to-bottom when possible, use quality checklist

#Scaffold commands

# Experiment notebook
uv run --python 3.12 python "$JUPYTER_NOTEBOOK_CLI" \
  --kind experiment \
  --title "Compare prompt variants" \
  --out output/jupyter-notebook/compare-prompt-variants.ipynb

# Tutorial notebook
uv run --python 3.12 python "$JUPYTER_NOTEBOOK_CLI" \
  --kind tutorial \
  --title "Intro to embeddings" \
  --out output/jupyter-notebook/intro-to-embeddings.ipynb

#Dependencies

Optional for local execution:

uv pip install jupyterlab ipykernel

The bundled scaffold script uses only the Python standard library.

This skill is from the OpenAI Skills Catalog.