red-team-tribunal

Multi-agent adversarial verification system. Use when you need thorough code review, security auditing, or quality validation. Spawns three specialized agents (Skeptic, User Proxy, Optimizer) that must reach consensus before code can be approved.