Codex plugin for structured task planning, adversarial plan verification, and evidence-based review. 3 skills, 0 subagents — adds durable planning discipline without replacing native Codex execution.
Predict which files AI coding agents will miss — implementation spike + compiler + reference graph
Adversarial plan verification for Claude Code
Verify runtime-generated agent plans before they run. Resolves minimal planner output against application-owned module contracts and refuses policy-breaking plans before any generated step executes. Optional backend for Maida.
To associate your repository with the plan-verification topic, visit your repo's landing page and select "manage topics."