Early access · 18 of 60 problems runnable
Design the algorithm before you write the code.
Build a flowchart that actually runs. Step through it node by node, watch every variable change, and test it against real cases — then carry the same logic into Python.
1. Read the problem
Every challenge comes with a typed input/output contract, constraints and worked examples.
2. Build a runnable flowchart
Connect Input, Assign, Decision, While and Return nodes. Expressions are parsed and validated as you type.
3. Run, step, submit
Execute on samples with a full variable trace, set breakpoints, then submit to the judge for hidden tests.
What works today
- • Integers, booleans, strings, arrays (including nested)
- • Python-style arithmetic: + − * / // % and comparisons
- • Branching, while-loops, early returns
- • Step-by-step trace, breakpoints, variable diff
- • Structural validation incl. “used before assigned” checks
- • Step and time limits that stop infinite loops
Problems that need sets, maps, stacks, queues, recursion or graph primitives are listed as Coming soon until the engine supports them.