marimo-studio 0.0.2

Design custom views for Marimo notebooks

Marimo Studio

Build focused dashboards, reports, and tools from a Marimo notebook. Keep calculations, reactive controls, plots, tables, and anywidgets in Python, then arrange the cells and values each audience needs in custom HTML and CSS views. One notebook can serve several views.

Try the example

From a checkout with uv installed, run:

uv run marimo-studio examples/analysis.py

The workspace opens with the notebook and finished dashboard side by side. Change the scenario or quarter and watch the metrics, summary, and table update from the same Python session. Split either pane when you want the HTML or CSS source in the workspace. Open Pane, choose Add Source, then place it on any side of the current pane.

The example pairs an ordinary notebook with one view under examples/__marimo__/studio/analysis/dashboard/.

Build a view

Open any notebook with Studio:

uvx marimo-studio analysis.py

Studio creates a blank dashboard view beside the notebook. Place complete cell outputs with <marimo-cell> and JSON-compatible Python values with mo-value. Controls and widgets remain connected to the notebook kernel. In edit mode, a control change in the editor or any attached preview updates the other open views.

Keep the editor and preview open while you or an agent edits the view. Saved HTML and CSS refresh around the running Python session. The CLI exposes notebook discovery and runtime validation as structured output:

uvx marimo-studio inspect analysis.py --display --format json
uvx marimo-studio check analysis.py --runtime --format json

Create another view from the same notebook:

uvx marimo-studio view add executive analysis.py

View source stays under __marimo__/studio/ with the notebook. Python remains the source of calculations and reactive behavior.

Run the view

Serve the default view through Marimo's server:

uv run --with marimo-studio \
  marimo run analysis.py \
  --sandbox \
  --headless

Visitors receive the custom view backed by a live Python session. Marimo keeps ownership of notebook execution, authentication, sessions, and server APIs.

Use Python 3.11 or newer with Marimo 0.23.16 or newer.

Documentation

• Create your first view
• Design a view
• Share a view
• CLI and configuration
• Python API