Issue del repository

Two thinking frameworks for AI agents that form a complete generate → verify loop. first-principles breaks analogical reasoning to find root causes before coding. adversarial-review launches parallel agents to attack your code from six dimensions before it hits production.