6-layer AI governance: safety gates, evidence-based debugging, anti-slack detection, and machine-enforced hooks. Makes AI safe, thorough, and honest.

六层AI治理：安全闸门、基于证据的调试、防懈怠检测和机器强制挂钩。让AI更安全、更严谨、更诚实。