design-validation
Skill0 starsChanged 10 days agoMITView on GitHub
Final design review for UI work — run the production checklist before considering any UI complete. Load when finishing or reviewing UI: verifying accessibility, contrast, browser reality, responsiveness, or answering "does this look good?". Companion to the design-principles skill, which carries the design rules; do not load this skill for creating or changing UI.
2 other projects use a version of this.
Adopted by
Most recently taken up by
- oaslananka/easyeda-mcp-protheir own version, first seen Nov 1, 2026
- curvedinf/dev-skillstheir own version, first seen Oct 20, 2026
Other versions2 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 2
- oaslananka/easyeda-mcp-pro★ 67same name
- curvedinf/dev-skills★ 7same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository34 other files it gives its agents.
- chatgpt/AGENTS.mdAGENTS.md
- opencode/AGENTS.mdAGENTS.md
- zcode/AGENTS.mdAGENTS.md
- claude/CLAUDE.mdCLAUDE.md
- agent-doc-authoringSkill
- agent-experienceSkill
- ai-image-promptsSkill
- api-behavior-flowSkill
- architecture-and-performanceSkill
- behavioral-testingSkill
- behavior-auditSkill
- browser-tabs-uiSkill
- code-qualitySkill
- context-compactSkill
- design-principlesSkill
- design-to-deliverySkill
- e2e-test-authoringSkill
- engineering-reviewSkill
- excalidrawSkill
- external-ai-consultSkill
- google-omni-videoSkill
- llm-agent-tool-disciplineSkill
- parallel-subagentsSkill
- qa-test-scenariosSkill
- reliable-executionSkill
- research-and-reuseSkill
- server-security-hardeningSkill
- session-handoffSkill
- shadcn-librarySkill
- sudo-handoffSkill
- task-intakeSkill
- tauri-v2-docsSkill
- test-failure-triageSkill
- user-behavior-flowSkill
The file
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 4 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 22
After completing UI work, before declaring it done.
- Very specific: line 26
Do not use for creating or changing UI — that is design-principles.
- Distinctive: line 30
Run the application (or a live preview) — code inspection alone is not verification.
- Distinctive: line 32
Prefer an existing authorized local preview.
- Distinctive: line 35
Run applicable checklist items.
- Very specific: line 36
Keep non-obvious fixes documented as design decisions (see design-principles → Design Decisions).
≈ “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
Design Validation — ZCode
Purpose
Turn design judgment into verifiable checks. The design-principles skill
decides how to design; this skill proves the result before it ships. If any
item below fails, fix it before considering the UI complete.
When to use
- After completing UI work, before declaring it done.
- Reviewing an existing UI ("does this look good?").
- Debugging visual or UX issues reported against shipped UI.
Do not use for creating or changing UI — that is design-principles.
Workflow
- Run the application (or a live preview) — code inspection alone is not verification. Prefer an existing authorized local preview; inspect startup scripts before launching. Do not trigger seeds/migrations or restart user processes just for visual review. If no safe preview is available, complete static review and state rendered checks as not run.
- Walk the affected flows twice: once with pointer, once with keyboard.
- Capture screenshots at the target breakpoints.
- Run applicable checklist items. For a review-only request, report findings without editing; for implementation work, fix in-scope defects. Mark irrelevant items N/A and unavailable checks unverified, never passed.
- Keep non-obvious fixes documented as design decisions (see
design-principles→ Design Decisions).
Final Design Review
Checklist
- Existing components reused, nothing re-invented
- Responsive behavior checked at desktop, tablet, and mobile
- RTL/LTR verified: layout direction and directional icons mirror correctly
- Keyboard navigation works and focus states are visible
- Text/non-text contrast checked in applicable states, with numeric evidence or a contrast checker; do not treat exempt inactive controls as automatically requiring normal active-text contrast
- No unnecessary decoration or animation
- Matches existing product identity and design system
Tickable checks — Visual
- Spacing comes from the existing tokens, no random margins or padding
- Colors come from the token system, no raw hex scattered in components
- One clear primary action per view unless the context requires multiple equivalent actions
- New components exist only where existing patterns do not solve the problem
Tickable checks — UX
- Primary action is obvious in the first viewport
- All component states implemented (default, hover, active, disabled, loading, error, empty)
- Product-level states present where relevant (first run, empty, loading skeleton, offline, syncing, permission denied)
- Mobile and RTL behavior tested, not assumed
Tickable checks — Quality
- No placeholder-looking UI or fake data
- No inconsistent components
- No generic AI style
- Smallest change that satisfies the requirement — nothing over-engineered
Accessibility verification
- Keyboard walkthrough passes: logical tab order, every interactive element reachable and operable
- Visible focus states on all interactive elements
- No color-only meaning: status also conveyed by icon, text, or pattern
- Icon-only controls have accessible names
- Touch targets ≥ 44×44 CSS px on touch interfaces
- No unintended focus traps; modal dialogs intentionally contain focus while open and restore it on close
prefers-reduced-motionhonored
Browser reality
- No console errors
- No layout overflow at target widths
- No inaccessible interactive elements (missing roles, names, or labels)
Visual verification
- Screenshots captured at target breakpoints (desktop, tablet, mobile)
- Screenshots reviewed for alignment, spacing, and hierarchy — not just code reviewed
- Result compared against the existing product language, not an imagined ideal
Common mistakes
- Trusting that correct code means correct visuals — run it and look.
- Checking the checklist only in the happy path; walk error and empty states.
- Skipping the visual comparison against the existing product.
Validation
Record the actual widths, themes/directions and flows checked, plus failures, N/A and checks not run. Review completion is not product sign-off; a review may complete with documented defects. Only claim visual implementation validation when the applicable rendered checks were actually observed.
The Final Design Review above is the validation: fix every failed item before finishing. A UI is complete only when the checklist passes in the running application.
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.

