Ui neuro ergo
Autonomous neuro-ergonomic + a11y UI audit and iteration skill. Combines WCAG 2.2 AA standards, cognitive science principles (Fitts Law, CLT, Gestalt), and a Chrome DevTools MCP screenshot loop for production-ready delivery. Triggers on neuro ergo, a11y, accessibility audit, cognitive load, UX review, WCAG, ergonomie, accessibilite.From its SKILL.md
npx -y skills add TheWatcher01/skills --skill ui-neuro-ergoAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 0 stars0 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
5.4 KB, ~1.3k tokens by cl100k_base, as published. Nobody here has run it
UI Neuro-Ergo Skill
Autonomous workflow for neuro-ergonomic + accessible UI work with visual iteration via Chrome DevTools MCP. Delivers production-ready results without requiring human intervention between iterations.
Overview
This skill combines:
- Neuro-ergonomics — cognitive science applied to UI: reduce friction, lower cognitive load, amplify signal
- Accessibility (WCAG 2.2 AA) — inclusive design, screen reader support, keyboard navigation, contrast
- Visual Iteration Loop — Chrome DevTools MCP screenshots for autonomous see→fix→verify cycles
When to load references
- Neuro-ergonomic principles → load
references/neuro-ergo-principles.md - WCAG 2.2 AA checklist → load
references/a11y-checklist.md - Chrome MCP iteration protocol → load
references/iteration-protocol.md
Workflow
Phase 1 — Audit
- Open Chrome DevTools MCP → navigate to the target page
- Take screenshot → capture current state (full page)
- Take a11y tree snapshot (
take_snapshot) → see ARIA roles, labels, structure - Read the source code → analyze component structure
- Run axe audit via
evaluate_script:() => { // Check for basic a11y violations visible without axe const issues = []; // Missing alt on images document .querySelectorAll("img:not([alt])") .forEach((el) => issues.push({ type: "img-no-alt", el: el.outerHTML.slice(0, 80) }), ); // Buttons without accessible name document .querySelectorAll("button:not([aria-label]):not([aria-labelledby])") .forEach((b) => { if (!b.textContent.trim()) issues.push({ type: "button-no-name", el: b.outerHTML.slice(0, 80), }); }); // Inputs without label document .querySelectorAll("input:not([aria-label]):not([id])") .forEach((i) => issues.push({ type: "input-no-label", el: i.outerHTML.slice(0, 80) }), ); // animate-pulse without reduced motion guard document .querySelectorAll('[class*="animate-"]') .forEach((el) => issues.push({ type: "animation-check", cls: el.className }), ); return issues; }; - Compile audit report with findings in 4 categories:
- 🔴 Critical (WCAG A violations, blocking keyboard/SR users)
- 🟠 Ergonomic (cognitive load, Fitts' Law, visual hierarchy)
- 🟡 WCAG AA (contrast, live regions, target sizes)
- 🟢 Polish (micro-interactions, motion, progressive disclosure)
Phase 2 — Plan Fixes
Load references/neuro-ergo-principles.md and references/a11y-checklist.md.
For each finding, define the specific fix with exact code changes.
Prioritize: Critical → WCAG AA → Ergonomic → Polish.
Phase 3 — Implement
Apply all changes to the source file(s). Stagger changes in logical groups if file is large:
- Structure & ARIA (landmarks, roles, labels, live regions)
- Keyboard & Focus (skip link, focus ring, tab order)
- Cognitive load (chunking, visual hierarchy, reduced noise)
- Motion & animation (
prefers-reduced-motion, remove unnecessary pulse) - Contrast & sizing (text contrast, touch targets min 44×44px)
Phase 4 — Visual Iteration
Reload the page and take before/after screenshots. Fix any regressions:
take_screenshot → analyze → find issues → fix code → navigate/reload → take_screenshot → ...
Iterate until:
- No critical a11y violations visible
- Visual hierarchy is clear (1 primary action per view, clear F-pattern scan path)
- All interactive elements have visible focus indicators
- Motion respects
prefers-reduced-motion - Skip navigation link present
- Live region announces filter results to screen readers
- Touch targets ≥ 44×44px on mobile viewport
- Headings form a logical outline (h1 → h2 → h3, no skips)
Phase 5 — Production Checklist
Before declaring done, verify:
- TypeScript compiles without errors (
pnpm buildor check editor diagnostics) - No arbitrary Tailwind values (
w-[...],text-[...]) — unless using CSS vars - All
<button>have accessible names - All
<input>have<label>oraria-label -
<article>elements havearia-labelledbypointing to their heading - Dynamic count updates use
role="status"oraria-live="polite" - Animations wrapped in
motion-safe:or@media (prefers-reduced-motion: no-preference) - Color is not the only differentiator (icon+color, not color alone)
Reference Files
Load these on demand:
references/neuro-ergo-principles.md— cognitive science principles (CLT, Fitts' Law, Gestalt, Miller's Law, F-pattern)references/a11y-checklist.md— WCAG 2.2 AA checklist + JS audit script for React/Next.jsreferences/react-a11y-patterns.md— copy-paste fix patterns (skip link, live region, article+aria-labelledby, radiogroup, motion guard, touch target)references/iteration-protocol.md— Chrome DevTools MCP tool sequence and iteration exit criteria
What ships with it: 4 files
24.3 KB alongside SKILL.md
references/
- a11y-checklist.md9.0 KB
- iteration-protocol.md6.1 KB
- neuro-ergo-principles.md4.6 KB
- react-a11y-patterns.md4.5 KB