Expert study blocks

Expert blocks apply professional review methods to your screens — no audience required. Findings carry a severity, confidence, and recommendation, and reference the screens they affect. Every expert block also accepts freeform Additional notes for the AI to take into account.

UX review

A structured review across focus areas. By default it covers all of them — clarity, hierarchy, navigation, content, trust, accessibility heuristics, and cross-screen consistency; untick All focus areas to narrow the review to the ones you're worried about. Cross-screen consistency is the one human reviewers most often skip — the AI checks that repeated elements behave and look the same everywhere.

Accessibility review

Reviews the screens for accessibility problems: contrast, target size, labelling, and focus order. Nothing to configure beyond optional notes — run it on everything, early. Treat the output as a screening pass, not a conformance audit: automated review of static screens can't replace testing with assistive technology.

Designer task walkthrough

Listed as Task walkthrough in the block picker. The same walkthrough as the audience version — Task or goal (required), optional Success criteria — but evaluated as an expert cognitive walkthrough rather than through a persona: at each step, is the action visible, is it clearly the right one, is the feedback unambiguous?

Heuristic evaluation

Evaluates the screens against Nielsen's 10 usability heuristics — visibility of system status, match with the real world, user control and freedom, consistency and standards, error prevention, recognition rather than recall, flexibility and efficiency, aesthetic and minimalist design, error recovery, and help & documentation. Run all ten (All heuristics) or select a subset; findings name the heuristic they violate.

Design comparison

Compares two or more variants on craft rather than audience appeal: Comparison goal (required) and at least two Screens to compare. The result assesses each option against the goal, with its tradeoffs. Use the audience version when the question is "which lands with users", this one when it's "which is the better design".