jupyter-notebooks — Codex skill: what it does, how to install
This skill builds, edits, and checks repeatable SQL or Python notebooks. Use it when an analysis must be readable, rerunnable, and easy to audit.
npx skills add openai/plugins --skill jupyter-notebooks -a codexWhat jupyter-notebooks does
The skill treats a Jupyter notebook as a reader-facing analysis artifact. It can shape reports, experiments, diagnostics, data-quality checks, and tutorials, while keeping data loading, preparation, calculations, charts, and interpretation in a clear sequence.
For existing files, it preserves the original intent and limits structural churn. Notebook-aware tools such as JupyterLab, nbformat, and nbclient are supported for setup and validation. SQL may stay in SQL cells or separate query files, while assumptions, parameters, and source references remain visible beside the work.
Validation runs every cell in order with python -m jupyter nbconvert --execute --to notebook --inplace path/to/notebook.ipynb. The notebook cannot be checked this way unless the execution environment has the required kernel, dependencies, and data.
What it helps with
- build a repeatable Jupyter notebook for analysis
- separate SQL, Python, data preparation, and charts
- run notebook cells in a consistent order
- keep sources, parameters, and assumptions with the analysis
- refactor an existing notebook with limited structural changes
Example prompts for the agent
- Create
analysis.ipynbfor a repeatable SQL and Python analysis, with clear sections and an execution check. - Run
analysis.ipynbfrom the first cell to the last, connect the conclusions to cell outputs, and list the steps needed to reproduce the check. - Edit
report.ipynbwhile preserving its purpose, then separate data preparation from charting.
How to install jupyter-notebooks
1Skills CLI (Vercel)
npx skills add openai/plugins --skill jupyter-notebooks -a codex2Manually: put the files in place
.agents/skills/jupyter-notebooks/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add openai/plugins --skill jupyter-notebooks -a claude-code2Manually: put the files in place
~/.claude/skills/jupyter-notebooks/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/jupyter-notebooks/
Manually: put the files in place
.cursor/skills/jupyter-notebooks/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install openai/plugins jupyter-notebooks2Manually: put the files in place
.github/skills/jupyter-notebooks/SKILL.mdPlatform CLI
gemini skills install https://github.com/openai/plugins --path plugins/data-analytics/skills/jupyter-notebooksManually: put the files in place
.devin/skills/jupyter-notebooks/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: Codex skills: how to install, where they live, how to call them
Questions about jupyter-notebooks
How do I install jupyter-notebooks in Codex?
Run npx skills add openai/plugins --skill jupyter-notebooks -a codex, then check .agents/skills/jupyter-notebooks/SKILL.md.
What does the jupyter-notebooks skill do?
It creates, edits, and validates SQL or Python Jupyter notebooks. Its workflow keeps the analysis readable, repeatable, and tied to executed cell results.
What is required to validate a notebook?
The execution environment needs the relevant data, dependencies, and kernel. Run python -m jupyter nbconvert --execute --to notebook --inplace path/to/notebook.ipynb to validate the file.
