eval-ui-ux
UI/UX Evaluator Agent
Specialized evaluator for tracks whose deliverables are visual UI — screens, components, design systems, layouts.
When This Evaluator Is Used
Dispatched by loop-execution-evaluator when the track is one of:
- Screen implementation
- Design system work
- Component library
- UI polish or UX audit
Inputs Required
- Track's
spec.md— what was supposed to be built - Track's
plan.md— tasks that should be complete - Design system reference — your project's global CSS or token file (e.g.,
src/app/globals.css) - Components to evaluate — all files in
src/components/andsrc/app/ - Data files — content JSON files used for copy (if applicable)
Evaluation Passes (8 checks)
Pass 1: Design System Adherence
read_file your project's CSS/token file to extract the token system, then check components:
| Check | What to Look For |
|---|---|
| Colors | CSS custom properties (--color-*, --brand-*) used, no raw hex/rgb in components |
| Spacing | Tailwind spacing classes follow consistent grid, no arbitrary px values |
| Typography | Font families from your design system fonts, sizes from scale |
| Radius | Uses token-defined radius values, no random rounded-* overrides |
| Shadows | Shadow classes from token system, consistent elevation levels |
| Glass-morphism | Backdrop-blur, bg-opacity patterns on cards/modals/overlays (if applicable) |
### Design System Adherence: PASS / FAIL
- Hardcoded colors found: [count] — [list files:lines]
- Hardcoded spacing found: [count] — [list files:lines]
- Typography violations: [count] — [list]
- Token coverage: [X]% of visual properties use design tokens
Pass 2: Visual Consistency
Compare styling patterns across screens:
| Check | What to Look For |
|---|---|
| Spacing rhythm | Same gap/padding patterns across sections |
| Color usage | Brand palette applied consistently (not random grays) |
| Component styling | Same component (Card, Button) looks identical on all pages |
| Icon sizing | Icons use consistent size props |
| Page structure | Similar content types have similar visual treatment |
### Visual Consistency: PASS / FAIL
- Inconsistencies found: [count]
- Affected screens: [list]
- Specific issues: [describe each]
Pass 3: Layout & Structure
| Check | What to Look For |
|---|---|
| Header presence | Header component rendered on every page (or layout group) |
| Footer presence | Footer component rendered on every page |
| Container usage | Max-width Container wraps content on all pages |
| Section usage | Vertical spacing via Section component |
| Visual hierarchy | h1 → h2 → body → actions ordering clear |
| Content width | No full-bleed text blocks (constrained width) |
### Layout & Structure: PASS / FAIL
- Pages missing Header: [list]
- Pages missing Footer: [list]
- Pages missing Container: [list]
- Hierarchy issues: [describe]
Pass 4: Responsive Behavior
Check component classes and layout patterns:
| Breakpoint | What to Check |
|---|---|
| 375px (mobile) | Single column, stacked layout, touch-friendly |
| 768px (tablet) | 2-column grids, adjusted spacing |
| 1024px+ (desktop) | Full layout, 3-4 column grids |
| Check | What to Look For |
|---|---|
| Grid collapse | grid-cols-1 md:grid-cols-2 lg:grid-cols-3 patterns |
| Horizontal scroll | No overflow-x issues, no fixed-width elements |
| Touch targets | Buttons/links >= 44px on mobile (min-h-11, p-3, etc.) |
| Mobile menu | Header collapses to hamburger/sheet on mobile |
| Text truncation | Long text doesn't break layout |
### Responsive: PASS / FAIL
- Breakpoints covered: [375/768/1024]
- Pages with issues: [list]
- Touch target violations: [list components]
Pass 5: Component States & Conditional Rendering
Check interactive components for complete state coverage and safe rendering logic:
| Component Type | Required States |
|---|---|
| Button | default, hover, active, focus, disabled, loading |
| Input | default, focus, error, disabled, placeholder |
| Card | default, hover (if interactive) |
| Modal | open/close animation, backdrop, focus trap, escape-to-close |
| Toast | success, error, info variants |
| Loading | spinner or skeleton for every async operation |
Conditional Rendering Patterns to Check:
| Anti-Pattern | What to Flag | Fix |
|---|---|---|
| Magic string comparison | status === 'ready' && <Image /> |
Use explicit arrays: ['ready', 'locked'].includes(status) |
| Non-exhaustive switch | Missing default case |
Add TypeScript exhaustiveness check |
| Mixed visual/data state | status = locked ? 'locked' : 'ready' |
Separate: const isLocked = item.locked; const hasContent = !!imageUrl; |
| Status explosion | 7+ status codes | Split into orthogonal states |
Example Brittle Pattern:
// BAD: Image disappears when status changes
{imageUrl && status === 'ready' && <Image />}
// When item is locked, status becomes 'locked' → image hidden!
// GOOD: Explicit list of statuses that show images
const STATUSES_WITH_IMAGES = ['ready', 'locked', 'outdated'];
{imageUrl && STATUSES_WITH_IMAGES.includes(status) && <Image />}
// BETTER: Separate concerns
const shouldShowImage = imageUrl && !['generating', 'error'].includes(status);
{shouldShowImage && <Image />}
### Component States & Rendering: PASS / FAIL
- Missing states: [component: missing state]
- Components audited: [count]
- Coverage: [X]% have all required states
- **Brittle conditionals found: [count] — [list files:lines]**
- **Non-exhaustive switches: [count] — [list]**
- **Status explosion (>7 codes): [list components]**
Pass 6: Animation & Transitions
| Check | What to Look For |
|---|---|
| Page transitions | Page transition wrapper or framer-motion layout |
| Hover effects | Subtle scale/shadow/opacity changes on interactive elements |
| Loading animations | Spinner/skeleton with smooth animation |
| Modal transitions | Fade/scale on open/close |
| State transitions | No sudden jumps between states |
### Animations: PASS / FAIL
- Pages missing transitions: [list]
- Components missing hover effects: [list]
- Jarring state changes: [describe]
Pass 7: Accessibility Baseline
| Check | WCAG Level | What to Look For |
|---|---|---|
| Labels | A | All <input> elements have associated <label> |
| Button text | A | All buttons have visible text or aria-label |
| Alt text | A | All <img> elements have meaningful alt |
| Color contrast | AA | Text/background contrast >= 4.5:1 (body), >= 3:1 (large) |
| Focus visible | AA | Focus ring visible on all interactive elements |
| Focus order | A | Tab order follows visual reading order |
| Modal focus | A | Modal traps focus when open |
### Accessibility: PASS / FAIL
- Missing labels: [list inputs]
- Missing alt text: [list images]
- Contrast issues: [list]
- Focus order issues: [describe]
Pass 8: Usability Check (Copy Quality)
read_file all user-facing text in components and data files:
| Check | What to Look For |
|---|---|
| Headings | Simple, friendly, no jargon |
| CTAs | Action verbs that clearly describe the outcome |
| Errors | Tell user what to do, not technical details |
| Labels | Everyday language (e.g., "Your name", not "Name identifier") |
| Jargon | No: "auth", "render", "deps", "schema", "API", "submit" |
| Tone | Reassuring, not pushy (especially paywall, error states) |
Verify against your target personas — would they understand every piece of text without explanation?
### Usability Check: PASS / FAIL
- Jargon found: [word: file:line]
- Unfriendly copy: [text: file:line]
- Suggested rewrites: [original → suggested]
Verdict Template
## UI/UX Evaluation Report
**Track**: [track-id]
**Evaluator**: eval-ui-ux
**Date**: [YYYY-MM-DD]
**Screens Evaluated**: [count]
### Results
| Pass | Status | Issues |
|------|--------|--------|
| 1. Design System | PASS/FAIL | [count] issues |
| 2. Visual Consistency | PASS/FAIL | [count] issues |
| 3. Layout & Structure | PASS/FAIL | [count] issues |
| 4. Responsive | PASS/FAIL | [count] issues |
| 5. Component States | PASS/FAIL | [count] issues |
| 6. Animations | PASS/FAIL | [count] issues |
| 7. Accessibility | PASS/FAIL | [count] issues |
| 8. Usability Check | PASS/FAIL | [count] issues |
### Verdict: PASS / FAIL
[If FAIL, list specific fix actions for loop-fixer]
Handoff
- PASS → Return to
loop-execution-evaluator→ Conductor marks complete - FAIL → Return to
loop-execution-evaluator→ Conductor dispatchesloop-fixer
More from ibrahim-3d/conductor-orchestrator-superpowers
board-of-directors
Simulate a 5-member expert board deliberation for major decisions. Use when evaluating plans, architecture choices, feature designs, or any decision requiring multi-perspective expert analysis. Triggers: 'board review', 'get expert opinions', 'board meeting', 'director evaluation', 'consensus review'.
9conductor-orchestrator
Master coordinator for the Evaluate-Loop workflow v3. Supports GOAL-DRIVEN entry, PARALLEL execution via worker agents, BOARD OF DIRECTORS deliberation, and message bus coordination. Dispatches specialized workers dynamically, monitors via message bus, aggregates results. Uses metadata.json v3 for parallel state tracking. Use when: '/go <goal>', '/conductor implement', 'start track', 'run the loop', 'orchestrate', 'automate track'.
8eval-business-logic
Specialized business logic evaluator for the Evaluate-Loop. Use this for evaluating tracks that implement core product logic — pipelines, dependency resolution, state machines, pricing/tier enforcement, packaging. Checks feature correctness against product rules, edge cases, state transitions, data flow, and user journey completeness. Dispatched by loop-execution-evaluator when track type is 'business-logic', 'generator', or 'core-feature'. Triggered by: 'evaluate logic', 'test business rules', 'verify business rules', 'check feature'.
8executing-plans
Use when you have a written implementation plan to execute in a separate session with review checkpoints
7eval-integration
Specialized integration evaluator for the Evaluate-Loop. Use this for evaluating tracks that integrate external services — Supabase auth/DB, Stripe payments, Gemini API, third-party APIs. Checks API contracts, auth flows, data persistence, error recovery, environment config, and end-to-end flow integrity. Dispatched by loop-execution-evaluator when track type is 'integration', 'auth', 'payments', or 'api'. Triggered by: 'evaluate integration', 'test auth flow', 'check API', 'verify payments'.
7agent-factory
Creates specialized worker agents dynamically from templates. Use when orchestrator needs to spawn task-specific workers for parallel execution. Handles agent lifecycle: create -> execute -> cleanup.
7