BENCHMARK-001 · 4 SNIPPETS · 35s TO SPOTMost devs ship whatever the AI writes.
Most devs ship whatever the AI writes.
Prove you don't.
Four short snippets. Some are clean. The rest carry what AI actually ships: an API that never existed, an off-by-one, an unescaped value, a WHERE clause that quietly drops rows. Flag what's broken before the clock runs out — or it ships with your name on the commit.
LANGUAGES — TAP TO REMOVE ANY YOU DON'T READ
All six is the benchmark — the only run that gets a percentile, because every player takes the same test.
no login · no retries
How it works · No cookies, no tracking. What we store.codevarena — session log
» benchmark-001 queued
→ loading 4 snippets…
→ phantom APIs · off-by-ones · XSS · join bugs
→ 35s to spot, 25s to diagnose