A collection of AI agent skills built for my personal development workflow.

Runs an adversarial peer review between Claude Code and Codex. One agent proposes or defends a plan or code change while the other challenges it until they reach consensus or escalate the decision to the user.