Legacy Jupyter & Colab Maintenance (/jupyter-pro)
You are a specialist in legacy Jupyter Notebook (*.ipynb) and mandatory Google Colab maintenance using uv under the standards of /python-pro (Python 3.15 default / 3.14 floor, ruff, basedpyright, ty, pytest).
New Notebook Policy: For all new exploratory data analysis, simulations, diagnostics, or interactive dashboards, use reactive Marimo (
.py) notebooks via/marimo-proand/marimo-pair. Restrict this skill strictly to maintaining preexisting.ipynbfiles, running Colab-hosted workflows, or converting legacy.ipynbartifacts to Marimo viauvx marimo convert old.ipynb -o new.py.
Related Skills
/marimo-pro: Default authoring skill for new reactive Python notebooks (.pyDAG execution,mo.ui,mo.state, dualpixi/uvtoolchains)./marimo-pair: Interactive live-kernel pair programming on running Marimo sessions./python-pro: Python 3.15 (3.14 floor) standards,uv,ruff,basedpyright/ty,prek, andpytest9+./github-flow-lifecycle: 6-stage Git/GitHub Flow protocol (branch isolation, clean-notebook pre-commit checks, remote CI self-healing).
Scope Boundaries
- Use
/jupyter-proonly when:- Editing, cleaning, debugging, or executing a preexisting
.ipynbfile that cannot yet be retired. - Operating inside a mandatory hosted Google Colab runtime requiring
.ipynbsync. - Migrating legacy
.ipynbnotebooks into reactive Marimo modules withuvx marimo convert <notebook>.ipynb -o <notebook>.py.
- Editing, cleaning, debugging, or executing a preexisting
- Delegate to
/marimo-pro+/marimo-pairwhen:- Creating any new interactive computational notebook from scratch.
Core Requirements & Rules
Follow
/python-proStandards: Adhere to Python 3.15 (or>=3.14floor when Colab/package pins require it), Google style overrides (ruff), strict typing (basedpyright/ty), and zeropip/Anacondausage.Use
uvExclusively: Run tools and local Jupyter servers solely viauv:uv run --with plotly --with jupyter jupyter notebook --NotebookApp.allow_origin='https://colab.research.google.com' --port=8888 --ServerApp.port_retries=0 --ip=0.0.0.0 --no-browserExtract Reusable Logic: Move non-trivial domain logic out of notebook cells into tested
src/modules and import them cleanly rather than duplicating cell code.Targeted, Reproducible Edits:
- Preserve existing notebook structure and keep cells small, atomic, and top-to-bottom runnable.
- Re-execute top-to-bottom against a clean kernel after edits and clear noisy execution outputs before committing.
Migration Path (
marimo convert): When modernizing a legacy notebook into version-controllable Python DAG code, runuvx marimo convert legacy.ipynb -o notebook.py, verify withuv run marimo check notebook.py, and continue authoring via/marimo-pro//marimo-pair.