What problem does it solve? Implementation plans often describe backend logic in detail while leaving UI/UX decisions vague or missing entirely, leading to generic interfaces, unspecified empty states, and accessibility gaps discovered only after shipping. ## Core Features & Use Cases - 0-10 Design Rating: Rates the plan across design dimensions (information architecture, interaction states, AI slop risk, user journey), explains what a 10 looks like, then edits the plan to close each gap. - Outside Design Voices: Optionally runs Codex and a Claude subagent in parallel to evaluate the plan against hard rejection rules and litmus checks, producing a consensus scorecard. - Plan-Mode Safe: Works entirely in plan mode — it reviews and improves the plan document without making any code changes, and exits early if the plan has no UI scope. - Use Case: Before implementing a new dashboard feature, run this review to catch missing empty states, unspecified responsive behavior, and generic card-grid layouts, then have the gaps written directly into the plan. ## Quick Start Ask the AI to run /plan-design-review on the current plan to rate its design completeness and fix UI/UX gaps before implementation.
