Notes on AI security, agent red teaming, and behavioral detection. Working out loud — failed attempts included.
Posts