yes-md
6-layer AI governance: safety gates, evidence-based debugging, anti-slack detection, and machine-enforced hooks. Makes AI safe, thorough, and honest.
The full SKILL.md could not be loaded right now. You can still install it from the upstream repo.