Mimic AI Security

An extension of Mimic's known-good enforcement model to AI-driven threats, blocking actions by AI agents, scripts, and automated processes that fall outside the authorized baseline at the kernel level.