Audit UX for Usability Issues

UX Audit is a skill that audits a target screen or flow for UX issues using Nielsen's 10 usability heuristics plus a mobile-specific UX checklist. Under Nielsen's heuristics it works through Visibility of System Status (loading states, success/error feedback, progress indicators), Match Between System and Real World (plain-language labels, recognizable icons, locale-appropriate formats), User Control and Freedom (back navigation, confirmation on destructive actions, undo, dismissible sheets, no dark patterns), Consistency and Standards (consistent action styling, color meaning, a 6px spacing grid), Error Prevention (distinct destructive buttons, blur-time validation, visible input constraints), Recognition Rather Than Recall (icon labels, visible current state, recent/frequent items), Flexibility and Efficiency (key actions reachable within 3 taps, minimum 44x44px touch targets, pull-to-refresh), Aesthetic and Minimalist Design (single primary task per screen, information hierarchy, card density limits), Help Users Recover from Errors (plain-language error messages with fix suggestions, isolated partial failures, retry on network errors), and Help and Documentation (guided empty states, onboarding, tooltips). It then runs a separate mobile-specific pass covering touch and gesture sizing, performance perception (skeleton screens within 300ms, optimistic updates, no layout shift), safe-area handling (notch, Dynamic Island, home indicator), and a dedicated dark-pattern check (no forced entry sheets, no exit-prevention dialogs, clear CTA labels).

Use it when you need to audit an existing screen against usability and mobile UX standards, taking the target as an argument. Do NOT use it for accessibility-only issues (use /ss-a11y instead), for design-system token or golden-rule compliance (/ss-review), for copy or microcopy quality specifically (/ss-copy), or for brand-new screens that don't exist yet - design those first with /ss-page or /ss-flow before auditing them.

Input is a target screen or flow, passed as the skill's $ARGUMENTS. Output is a structured report in five parts: a letter score from A+ to F with a breakdown, a list of Critical Issues that block usability and must be fixed, Major Issues that degrade the experience and should be fixed, Minor Issues that are polish-level nice-to-fixes, and Recommendations giving specific code changes for each issue found.

This skill is a community contribution sourced from bitjaru/styleseed's ss-audit skill (engine/.claude/skills/ss-audit), added to the collection on 2026-07-01 and licensed under MIT. Some of its heuristic checks reference conventions from that source project specifically - a Lucide icon set and a 5-level grayscale text hierarchy - so results are most directly applicable to codebases following similar design-system conventions, though the underlying Nielsen heuristics and mobile checks generalize broadly.

Frontend and product engineers who want a structured, checklist-driven UX review of a screen before shipping it - flagging missing loading or error states, inconsistent styling, undersized touch targets, and dark patterns - and who want the output pre-organized into a fix-priority order (critical/major/minor) with concrete recommendations rather than open-ended feedback.