Show HN: Cerno – CAPTCHA that targets LLM reasoning, not human biology
Summary
Cerno is an open-source CAPTCHA that uses motor-control analysis of maze interactions and 12 behavioral features to distinguish LLM-based bots from humans. It combines a growing-tree maze, adaptive proof-of-work, and server-side verification with an open TypeScript SDK. The project emphasizes trustless validation through seed-based maze regeneration and ephemeral ECDSA keys.