Rubric vs Vibes
Same screen, two passes, two very different issue counts.
the same screen judged twice by the same person produces two different issue counts — the rubric literally lights up blind spots the freeform pass walked straight past
What goes in, what comes out
Learner critiques the same flawed UI screen twice — first freeform, then guided by a structured rubric covering hierarchy, contrast, affordance, consistency, and feedback. The engine tags every rubric-relevant issue actually present in the mock and compares it to what each pass caught.
write freeform critique notes; then work the rubric checklist against the same screen
a coverage chart of issues caught freeform versus issues caught only once the rubric applied, highlighted directly on the mock
mock picker (5 seeded flawed screens); rubric dimension count (3 quick / 6 full); forced freeform-first toggle
free design-critique glossary page + Module 'Judging AI Output Like a Critic' L3 demo
Concepts click when you open the machinery.
Three labs are already live and free — the same hands-on style this playground brings to its module.