5 Days of AI Safety Engineering

Safety as architecture rather than a moderation setting. Threat models, agent identity and scope, untrusted input, tool abuse and blast radius, and data exfiltration controls.