Write reactive marimo notebooks as Python files with correct cell structure.
npx skills add https://github.com/mathisdrn/orca --skill marimo-notebook-mathisdrn
Please help me install this Agent Skill. Skill: marimo-notebook Source: https://github.com/mathisdrn/orca/tree/main/.agents/skills/marimo-notebook Command: npx skills add https://github.com/mathisdrn/orca --skill marimo-notebook-mathisdrn
@app.cell decorators, proper input/output variable flow, and PEP 723 inline dependency metadata. - Reactivity & State Guidance: Enforces marimo's DAG-based reactivity model, avoiding untracked mutations and unnecessary mo.state() usage. - Validation & Testing: Runs marimo check as a linter and supports pytest-based test cells inside notebooks. - Use Case: Ask for an interactive data exploration notebook with sliders and a plot; the Skill produces a runnable .py notebook that works both interactively via marimo edit and as a script via uv run. ## Quick Start Create a marimo notebook that loads a CSV, shows a filterable table, and plots the filtered data with a slider-controlled threshold.High-intent search queries and answers about installing and using this skill.
Define an `app = marimo.App()` and write each cell as a function decorated with `@app.cell`. Function parameters are cell inputs and returned variables are outputs; marimo builds the dependency graph automatically. Add PEP 723 script metadata at the top for dependencies.
Run `uv run notebook.py` for non-interactive script execution, `uv run marimo run notebook.py` to serve it as an interactive app, or `uv run marimo edit notebook.py` to edit it in the browser. Use `uvx marimo check notebook.py` to lint for common mistakes.
marimo stores notebooks as plain Python files instead of JSON and uses a reactive execution model where cells re-run automatically when their dependencies change. Jupyter executes cells manually in arbitrary order, which can leave hidden stale state.
marimo only renders the final expression of a cell. Indented or conditional expressions inside `if` blocks will not display; assign the result conditionally and make it the last expression instead.
Yes. Add `pytest` as a dependency and create cells containing only test code—functions starting with `test_` or classes starting with `Test`. Run `pytest notebook.py` from the command line to execute them.
Use `mo.state()` only for accumulated callback state or bidirectional sync between UI elements. For most cases, regular variables between cells and widget `.value` reads provide sufficient reactivity without explicit state management.
daidongCreate data visualizations from DataFrames using seaborn and matplotlib.
panicPaulConvert Streamlit apps into equivalent marimo notebooks by mapping widgets and layouts.
AubreyHanOrganize and validate Jupyter notebook workflows for data analysis and BigQuery investigations.
legoutConvert Jupyter .ipynb files into reactive marimo Python scripts.