chaseai-yt/grill-me-codex

A dual-model planning skill where Claude interrogates you to lock down a plan (Act 1), then OpenAI Codex adversarially reviews it over multiple rounds (Act 2), ensuring cross-provider validation before any code is written.

A two-act planning skill where Claude interrogates you to lock down the plan (Act 1), then hands it to OpenAI Codex for adversarial cross-provider review over multiple rounds until both models sign off (Act 2). Built on the grill-me skill with an added cross-model verification loop.

A Claude Code skill family that hardens implementation plans using two AI models. Act 1 grills the user to lock the plan, then Act 2 hands it to OpenAI Codex for adversarial cross-provider review over multiple rounds before any code is written.