What problem does it solve? Reviewers and CI systems produce feedback that is sometimes wrong, stale, or a poor fit for the codebase, and blindly complying leads to regressions and performative agreement. This Skill enforces a disciplined workflow for handling incoming review feedback: read everything first, restate each item, verify claims against the actual code, judge fit for the project, then fix or push back with evidence. ## Core Features & Use Cases - Six-phase response workflow: Read fully, restate, verify against code, judge fit, respond, and implement one item at a time. - Evidence-based pushback: Reject incorrect or out-of-scope suggestions with specific reasons such as file:line references, counter-examples, or violated conventions. - Routing to follow-up workflows: Convert feedback into actionable TODOs routed to bug diagnosis, test-driven development, or verification gates. - Use Case: A pull request receives five review comments and a failing CI check. Use this Skill to verify each claim against the diff, accept the two valid items, push back on one stale comment with evidence, and fix accepted items one at a time with verification after each change. ## Quick Start Invoke /review-response when PR review comments or CI failures arrive to process each item with verification before responding.
