planloop

Use when a written implementation plan, judicial draft plan, content plan, or any structured pre-execution document exists and needs independent verification before execution begins. Activates on user requests like "review the plan", "run a reviewer subagent", "verify this plan", "let's review before executing", "audit this plan", or "another review round". Especially required for large plans, plans with many constraints, plans involving AI-assisted judicial output subject to CNJ Resolução 615/2025 audit obligations, refactors touching multiple subsystems, AWS infrastructure changes, or any plan that will be executed by subagents. Always trigger this skill when the user mentions independent review, plan audit, plan verification, finding disposition, or wants to enforce a quality gate between plan-writing and execution. Do not use this skill for creating the initial plan, reviewing already-implemented code, or replacing the inline self-review inside writing-plans.