From preflight-ux-review
Use when reviewing a product surface, spec, screenshot, prototype, PR, or launch flow to catch UX risks before launch. Produces structured product-risk hypotheses, not user research claims.
How this skill is triggered — by the user, by Claude, or both
Slash command
/preflight-ux-review:catch-ux-risksThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Use this skill to review a product surface before launch.
Use this skill to review a product surface before launch.
The output is a product-risk artifact, not validated user research. Do not invent user quotes, survey findings, or behavioral certainty.
If the surface is thin, state the evidence limitation and keep confidence lower.
If the current workspace is a Preflight UX checkout, or PREFLIGHT_UX_REPO
points to one, prefer the repository artifacts over freeform review:
python3 scripts/preflight_ux.py locate.python3 scripts/preflight_ux.py validate when the task depends on repo
artifacts.panels/default.yaml or a supplied panel file for review lenses.taxonomy/issue-classes.yaml for issue-class names.schemas/report-finding.schema.json for normalized finding shape.python3 scripts/preflight_ux.py uxpanel run --surface <surface.md> --surface-type spec --panel panels/default.yaml --run-id <slug>.calibrate-ux-findings skill after findings are produced.If no checkout is available, still perform the review using the structure below and mention that repo validation was not run.
Look for concrete risks in these classes:
Return 5-10 findings as a table:
| Priority | Issue class | Risk hypothesis | Surface evidence | Recommendation | Confidence |
|---|
Then add:
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub sparckix/preflight-ux --plugin preflight-ux-review