usability-heuristic-reviewlisted

# usability-heuristic-review — heuristic evaluation (CT-UT) Reference: `fixture/expense-demo-review.md` (3 real violations found and confirmed against `examples/expense-demo/app/public/`, no fabrication). This skill covers the ISTQB CT-UT (usability testing) syllabus — the one test type QAIA's other skills leave entirely uncovered. **What this is not**: not a quantitative user survey (SUS score, task-completion-rate study with real users) and not A/B testing or eye-tracking — those need real users QAIA cannot recruit. This skill is the **formative, expert-review** half of CT-UT (heuristic evaluation + cognitive walkthrough), the half a single reviewer session-bound to a self-hosted app can actually do honestly. ## Steps 1. **Screen inventory.** For each key screen the test book covers (and any the user names), navigate and capture a snapshot/screenshot — same navigation discipline as `visual-check`/`a11y-audit`. 2. **Heuristic evaluation.** For each screen, check it against Nielsen's 10 heuristics (nngroup.com/articles/ten-usability-heuristics, verified against the source, not recalled): 1. Visibility of system status — is there feedback for every action, especially async ones (loading/pending state, not just a final result)? 2. Match between system and the real world — plain language, no internal jargon/error codes shown raw to the user. 3. User control and freedom — an obvious way out of an unwanted state (cancel, undo, back) without