agentsclimarketplace

Design review

Skill tansuasici/claude-code-kit/.claude/skills/design-review

Review UI implementation for design consistency, visual quality, AI-slop detection, and responsive behavior. Use when evaluating how a built UI looks and feels. For accessibility use /accessibility-audit; to build new components use /ui-component-builder.From its SKILL.md

Install
npx -y skills add tansuasici/claude-code-kit --skill design-review

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

6.5 KB, as published. Nobody here has run it

Design Review

Core Rule

Compare implementation to the design source of truth. Flag divergence with screenshots and component refs; never auto-apply visual fixes.

When to Use

Invoke with /design-review when:

  • After implementing UI changes and want to verify visual quality
  • Checking for design inconsistencies across the application
  • Detecting AI-generated design patterns ("AI slop") that look generic
  • Reviewing responsive behavior across viewport sizes
  • Before shipping user-facing UI changes

Default Behavior

When the user asks to audit, scan, review, or "give me a report" for design consistency / UI, produce the full design-review report automatically using the Process and Output Format sections below. Do not require the user to specify fields.

Only modify files when the user explicitly requests implement / fix / apply / refactor. By default, this skill is report-only.

Process

Phase 1: Inventory (first-pass leads)

This pass produces candidates, not findings. Treat counts as leads for deeper inspection in later phases. Do not report Phase 1 raw output as the final result.

Identify the project's design foundations:

  1. Check for DESIGN.md — if it exists, use it as the single source of truth for design tokens, colors, typography, spacing, and component styles
  2. Read style config — Tailwind config, CSS variables, theme files, design tokens
  3. Identify patterns — component library in use (shadcn, MUI, Chakra, custom)
  4. Check for design tokens — colors, spacing, typography, border-radius, shadows
  5. Cross-reference — compare style config against DESIGN.md (if present) and flag any deviations
  6. Note inconsistencies — multiple sources of truth, conflicting tokens

Phase 2: Visual Consistency Audit

Check for design coherence:

Spacing & Layout

  • Consistent spacing scale (4px/8px grid, Tailwind spacing)
  • Alignment of elements within and across pages
  • Consistent padding/margin patterns for similar components
  • Proper whitespace — neither cramped nor wasteful

Typography

  • Font scale is hierarchical and consistent
  • Line heights appropriate for each text size
  • Font weights used purposefully (not randomly bold)
  • Text colors have sufficient contrast
  • No more than 2-3 font families

Color

  • Color palette is cohesive (not random hex values)
  • Semantic color usage (success=green, error=red, warning=yellow)
  • Sufficient contrast ratios (WCAG AA: 4.5:1 normal text, 3:1 large text)
  • Dark mode consistency (if applicable)
  • No color-only information (icons/text supplement color)

Components

  • Buttons have consistent sizing, padding, and border-radius
  • Form inputs follow a uniform style
  • Cards/containers have consistent shadows and borders
  • Icons are from the same family and consistently sized
  • Interactive states (hover, focus, active, disabled) are uniform

Phase 3: AI Slop Detection

Flag generic AI-generated design patterns that look templated:

Common AI Slop Patterns

  • Uniform border-radius on everything (all rounded-xl or all rounded-lg)
  • Generic 3-column grid layouts with identical cards
  • Overuse of gradients, especially blue-to-purple
  • Drop shadows on every element
  • Generic stock-photo hero sections
  • "Dashboard" layouts that look like every AI demo
  • Excessive use of emojis as icons
  • Perfectly symmetric layouts with no visual hierarchy
  • Cookie-cutter component styling (every card identical)

How to Fix

  • Vary border-radius based on component size and context
  • Create visual hierarchy through size, weight, and spacing variation
  • Use color purposefully, not decoratively
  • Add intentional asymmetry and visual rhythm
  • Replace generic elements with purposeful, context-specific design

Phase 4: Responsive Behavior

Check responsive design across breakpoints:

  • Mobile (320-480px) — single column, touch targets ≥44px, no horizontal scroll
  • Tablet (768-1024px) — appropriate layout adaptation, not just scaled-down desktop
  • Desktop (1280+) — proper use of space, readable line lengths (45-75 chars)
  • Edge cases — very small screens (320px), very large screens (2560px+)
  • Content overflow — long text, many items, empty states

Phase 5: Interaction Quality

Review interactive elements:

  • Loading states — skeleton screens or spinners, not blank spaces
  • Empty states — helpful message and action, not blank page
  • Error states — clear, actionable error messages
  • Transitions — smooth, purposeful animations (not jarring or excessive)
  • Feedback — visual response to user actions (button press, form submit)

Output Format

# Design Review Report

## Design System Health
- Tokens defined: Yes/Partial/No
- Consistency score: High/Medium/Low
- Component library: [name or custom]

## Issues Found

### Critical (Breaks usability)
| # | Category | Location | Issue | Fix |
|---|----------|----------|-------|-----|
| 1 | Contrast | header.tsx:12 | Text contrast 2.1:1 | Use text-gray-900 instead of text-gray-400 |

### Major (Hurts quality)
| # | Category | Location | Issue | Fix |
|---|----------|----------|-------|-----|

### AI Slop (Looks generic)
| # | Pattern | Location | Fix |
|---|---------|----------|-----|
| 1 | Uniform border-radius | All cards use rounded-xl | Vary: rounded-lg for cards, rounded-md for buttons |

### Minor (Polish)
| # | Category | Location | Issue | Fix |
|---|----------|----------|-------|-----|

## Responsive Check
| Breakpoint | Status | Issues |
|------------|--------|--------|
| Mobile (375px) | Pass/Fail | ... |
| Tablet (768px) | Pass/Fail | ... |
| Desktop (1280px) | Pass/Fail | ... |

## Positive Patterns
[Well-implemented design decisions worth preserving]

Notes

  • This review focuses on implementation quality, not design direction — it won't redesign your UI
  • AI slop detection is about identifying patterns that look unintentionally generic, not banning specific styles
  • If the project has a DESIGN.md, compare implementation against its defined tokens, colors, and component styles
  • If the project has a Figma/design spec, compare implementation against it
  • For accessibility-specific review, use /accessibility-audit which covers WCAG compliance in depth
  • Requires reading the actual component/template code — not just CSS

Keep looking

Skills are one crate of 627,801. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.