Spark & Anvil Privacy

Ages 15–18 · Grades 9–12 · Computer Science

ReviewForge

Auto says it’s done. Find the test that proves it isn’t.

Soon an assistant writes the first draft of the code. The skill that matters is knowing whether it’s right. Here “Auto” — a fixed, rule-based assistant, not a live model — submits a function against a spec. Your job: pick the test that would actually catch its bug (the edge case, not the happy path), run it, find the faulty line, request changes, and confirm the fix. Auto never hands you the answer — you verify it.

The code is a teaching subset of the Swift language. Auto is deterministic — no live model, nothing that names you leaves your device.

Loading the review lab…

A case · you decide

Priya reviews the assistant's discount code

Priya, 17 — reviewing a function an AI assistant wrote for her school store's checkout

The idea in play: a reviewer does not trust that code works — they pick the test that would catch a bug, find the faulty line, and ask for the change.

The spec says: 10% off orders of $20 or more. The assistant's code is tidy, has a comment saying 'applies the discount', and passes the one test it came with: a $50 order gets $5 off.

One passing test on an easy case says very little. Bugs live at the edges.

The interesting order is exactly $20: the spec says it gets the discount.

Which test should Priya add first?

She has found the line: `if total > 20`. What does she do?

Review a change →

What next? Pick a door

Chosen from what this app is about. What you open is remembered on this device only.