ui-quality-verification
Skill1 starsChanged 8 days agoMITView on GitHub
Audit a UI for visual, responsive, keyboard, accessibility, and state quality using browser evidence. Use for planned audits or post-change checks, not for diagnosing an unexplained browser failure.
In this repository16 other files it gives its agents.
- AGENTS.mdAGENTS.md
- adaptive-workflowSkill
- architecture-mapSkill
- browser-diagnosticsSkill
- domain-modelingSkill
- feature-discoverySkill
- lean-senior-engineeringSkill
- release-readinessSkill
- research-plan-implementSkill
- skill-security-guardSkill
- spec-to-code-verificationSkill
- systematic-debuggingSkill
- transaction-safe-engineeringSkill
- ui-design-intelligenceSkill
- ui-implementationSkill
- verification-before-completionSkill
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 8
Verify the interface as a user experiences it, not merely as a successful build.
- Distinctive: line 20
Use browser evidence and the project's existing automation where available:
- Distinctive: line 24
Responsive: verify reflow, touch target size, readable tables/cards, and absence of unintended horizontal overflow.
- Distinctive: line 30
Accessibility: check names, roles, labels, error associations, status announcements, and color-independent meaning.
- Distinctive: line 44
If implementation is requested, make the smallest fix and rerun the affected visual, keyboard, responsive, and state checks.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
UI Quality Verification
Verify the interface as a user experiences it, not merely as a successful build. Use a development or authorized test environment with representative data.
Establish the audit surface
Confirm the URL, routes, viewports, user role, data setup, and whether the request is a
planned audit or a post-change check. Capture the current state before judging it.
For an unexplained failure, use browser-diagnostics; do not turn failure investigation
into a broad design audit.
Check the interface
Use browser evidence and the project's existing automation where available:
- Visual: compare screenshots at representative desktop, tablet, and mobile widths; inspect hierarchy, spacing, clipping, overlap, contrast, and focus treatment.
- Responsive: verify reflow, touch target size, readable tables/cards, and absence of unintended horizontal overflow.
- Keyboard: traverse interactive elements with Tab/Shift+Tab, activate controls with Enter/Space, verify visible focus and a logical order, and check dialogs/filters.
- States: exercise empty, loading, error, success, disabled, and permission-limited states. Confirm messages identify the next action.
- Accessibility: check names, roles, labels, error associations, status announcements, and color-independent meaning.
- Data journeys: for approvals, reconciliation, and financial data, verify totals, selected records, confirmation, and audit evidence.
For enterprise profiles, use the selected checks in
references/verification-checklist.md.
Record evidence and limits
Save screenshots and a concise findings list. Report the viewport, browser, route, data fixture, and checks actually run. Distinguish visual/interaction evidence from backend correctness. Do not claim accessibility compliance from a few automated checks.
If implementation is requested, make the smallest fix and rerun the affected visual, keyboard, responsive, and state checks. A successful build alone is not completion.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

