What problem does it solve? Plans handed off between agents often contain ambiguity, hidden assumptions, and choice-language that force the executing agent to guess intent. This Skill reviews a PLAN.md as a strict execution contract and blocks plans that are not safe for blind handoff. ## Core Features & Use Cases - Multi-pass plan review: Runs ambiguity, self-sufficiency, no-choice-language, standards, artifact hygiene, ownership-and-reuse, compatibility-policy, final-state-authoring, and handoff-readiness passes against the plan text. - Structured findings output: Writes a PLAN_REVIEW.md containing a ready/not_ready verdict, compact P# findings ordered by severity, required changes, and preserve-notes. - Initiative child packet checks: Verifies child packets are not placeholder plans and align with the parent initiative, imported context, and contract decisions. - Use Case: Before dispatching a task packet to an implementer agent, run this review to catch vague validation steps like "test it", unnamed owners, or unjustified backward-compatibility shims, then send the findings back to the requesting lead. ## Quick Start Review the PLAN.md in the current task packet as a blind-handoff execution contract and write the verdict and P# findings to PLAN_REVIEW.md.
