What problem does it solve? AI-generated interfaces tend to converge on generic template looks — default sans-serif type, safe blue accents, centered heroes with three feature cards — producing pages that could belong to any product. This Skill enforces a disciplined design process so every built page, component, or artifact is visually distinctive and grounded in its subject. ## Core Features & Use Cases - Seven-phase gated workflow: load taste values, pin the subject, write a design plan, run an anti-generic check, choose tooling, build, then pass restraint and critique reviews before shipping. - Anti-generic detection: checks plans against four known AI default looks (including the generic-minimal SaaS trap) using swap and regenerate tests before any code is written. - Tooling routing: directs work to web-artifacts-builder, artifact-design, or the dataviz skill depending on whether the output is a claude.ai artifact, repo code, or contains charts. - Use Case: When asked to build a portfolio case-study page for an AI chatbot product, the Skill produces a design plan with a subject-specific accent palette, characterful type pairing, and a live-conversation signature element — then builds and critiques it to a PASS verdict. ## Quick Start Ask the agent to build a landing page or redesign a screen for your product, and it will plan the design, check it against generic defaults, and build the interface.
