Teach your AI design judgment.
Not just components.

The gate now has eyes. /ss-score reads the code — but some of the worst 'AI-made' tells only exist in pixels: a hero that doesn't actually dominate, a lower third of dead whitespace, a web font that silently failed to load and fell back to Times, two colors that look like two accents once rendered, an empty state that's a blank white void. A human sees these in half a second; a code-reading gate misses all of them. New /ss-verify closes that gap: it renders the screen, screenshots it, and the agent LOOKS at the image — scoring the same StyleSeed gate against what it actually sees, including the empty/loading/error states, then fixes and re-renders until it passes. Code-clean is necessary; pixel-clean is the real bar. It never claims a visual pass without a real screenshot. Runs after /ss-score, and /ss-build finishes a renderable screen with it.