Drag blocks, wire functions, run and debug step-by-step. No loops in code — use visual loop groups instead. Pure dataflow.
Write Python functions, drag them as blocks on an infinite canvas. Connect outputs to inputs with wires. ELK auto-layout keeps it clean.
No for/while in code. Create loop groups, connect For (range) or While (condition) wires. Feedback wires + init for accumulation.
Run once, browse every step. Watch specific ports, set breakpoints with conditions. Step forward/back instantly — zero re-execution.
Inspect any value with a lazy tree view. Search keys/values, paginate huge lists, go to any index. Strings load in chunks.
Numeric data auto-charts. DataFrames as tables. 2D grids as heatmaps. All downsampled for performance.
All variables persist after Run — the pipeline IS its own trace. Watch only what you need. Clean code = max speed.