The SciStudio window

This page is a map of the SciStudio window: what each region is for and where to find things. For hands-on teaching — building a workflow step by step — use the tutorials in the Learning Center (learning-center.md) instead; this page tells you what the regions of the window are.

The layout

The desktop window has five regions, each described below: the toolbar along the top, the left rail with its sidebar, the stage in the center (with the bottom tab bar beneath it), and the preview column on the right.

In the AI browser layout (when SciStudio runs for an external AI app), the same regions rearrange: the icon rail and sidebar move to the right side of the window, and the preview column disappears — the Preview card inside the sidebar plays its role instead.

The toolbar

The bar along the top of the window:

Control What it does
Project menu The open project; switch between recent projects here.
New Creates a new project item: workflow, custom block, data type, MiniApp, note, plot, or package.
Import Bring an existing workflow or file into the project.
Save Save the current workflow or file.
Run Run the workflow on the canvas; shows Running while it executes.
Bring in my work Import a Jupyter notebook or past analysis and have the AI turn it into a workflow.
Packages Install and update block packages.
Learning Center Tutorials, tips, and the user guide's Reading tab.
Layout toggle Switch between the desktop and AI browser layouts.

The left rail and the sidebar

The narrow icon rail at the far left never resizes and stays visible even when the sidebar is collapsed — clicking an icon selects what the sidebar next to it shows. You can drag the icons (or press Alt+ArrowUp / Alt+ArrowDown on a focused icon) to reorder them; the order is remembered on this computer.

Rail icon The sidebar shows
Blocks The block palette — every block you can drag onto the canvas, with search and category chips.
Workflows The project's workflow files; the one on the canvas is highlighted.
MiniApps Your MiniApps — interactive data explorers you can open on a block's output or create with the AI assistant (see miniapps.md).
Data types The type catalogue — the data types known to this project and their fields.
Data The project's data folders (raw, processed, zarr, parquet, artifacts, exchange); double-clicking a file opens a preview tab.
Project The whole project as a file tree.

The sidebar is resizable and collapsible; collapse it when you want the canvas full width.

The stage: tabs, canvas, and MiniApps

The center of the window is the stage:

  • The tab bar above it holds one tab per open workflow (plus file tabs for code and previews), so you can keep several workflows open side by side.
  • The canvas shows the active workflow as a graph of blocks and wires. Selecting a block shows its settings in the Config tab of the bottom panel.
  • MiniApps run in their own layer over the stage: an open MiniApp keeps running (its Python included) until you close its tab, whether or not it is the one on screen.

The bottom tab bar

The panel along the bottom of the stage holds tabs for the workflow-wide surfaces. It is resizable, collapsible to just its tab strip, and can be pinned so it stays open.

Tab What it shows
AI Chat The embedded assistant (desktop layout; hidden in the AI browser layout, where the AI app's conversation plays this role).
Config The selected block's parameters and settings.
Logs The run log; the tab badge counts unread entries while you work elsewhere.
Terminal A terminal session inside the project.
Plots The project's plot cards.
History Run history — what ran, with which parameters, and restore points.
Git Branches, commits, and merges for the project's version control.

The preview column

The narrow column on the right exists only in the desktop layout.

Previewing data

Click any port on the canvas — input or output, before or after a run — and its preview opens here. The preview reads only a bounded sample of the object, never the whole thing, so even a very large dataset previews instantly. What you see depends on the data's type: a table renders as a scrollable table, an image as an image viewer, a spectrum as a line plot, and a package type uses the view its package ships. To make your own quick figures instead, see writing-plots.md; for interactive exploration, see miniapps.md.

Where to go next