From buyoung-skills
Reviews UI code or screenshots for UX guideline violations using 99+ rules across navigation, layout, accessibility, forms, performance, typography. Generates severity-rated fixes and guidance for web, mobile, desktop platforms.
npx claudepluginhub buyoung/skills --plugin backend-skillsThis skill uses the workspace's default tool permissions.
Comprehensive UX review and implementation guidance system based on 99+ validated guidelines across 8 domains.
Creates isolated Git worktrees for feature branches with prioritized directory selection, gitignore safety checks, auto project setup for Node/Python/Rust/Go, and baseline verification.
Executes implementation plans in current session by dispatching fresh subagents per independent task, with two-stage reviews: spec compliance then code quality.
Dispatches parallel agents to independently tackle 2+ tasks like separate test failures or subsystems without shared state or dependencies.
Comprehensive UX review and implementation guidance system based on 99+ validated guidelines across 8 domains.
Analyzes existing UI code or screenshots against the full guideline set to identify violations and improvement opportunities.
Provides proactive implementation guidance when building new UI components or features.
| Level | Criteria | Action |
|---|---|---|
| High | Blocks usability, causes accessibility failures, or breaks core interactions (e.g., missing focus states, no touch targets, color contrast violations) | Must fix before release |
| Medium | Degrades experience but has workarounds (e.g., missing hover states, suboptimal animation timing, no inline validation) | Should fix in current cycle |
| Low | Polish and best-practice improvements (e.g., easing functions, date formatting, breadcrumbs on shallow sites) | Fix when convenient |
| Platform | Scope |
|---|---|
| Web | Desktop browsers, responsive web, PWA |
| Mobile | iOS, Android native and hybrid apps |
| Desktop | Electron, Tauri, native desktop GUI |
| VisionOS | Spatial UI, gaze interaction |
| All | Cross-platform universal guidelines |
| Domain | Guidelines | Reference |
|---|---|---|
| Navigation & Motion | 14 | navigation-motion.md |
| Layout & Responsive | 15 | layout-responsive.md |
| Interaction & Touch | 14 | interaction-touch.md |
| Accessibility | 11 | accessibility.md |
| Forms | 10 | forms.md |
| Performance | 8 | performance.md |
| Visual & Feedback | 16 | visual-feedback.md |
| Emerging Patterns | 12 | emerging-patterns.md |
| Field | Required | Description |
|---|---|---|
| Target | Yes | UI code, component, page, or screenshot to review/guide |
| Mode | No | review (audit existing) or guide (advise new). Default: inferred from context |
| Platform | No | Web, Mobile, Desktop, VisionOS, or All. Default: inferred from code |
| Focus areas | No | Specific domains to prioritize (e.g., accessibility, forms, performance) |
| Scope | No | Full audit or targeted check. Default: full |
## UX Review: {Target}
### Summary
- **Platform**: {detected platform}
- **Total findings**: {count}
- **High**: {count} | **Medium**: {count} | **Low**: {count}
### Findings
#### [{Severity}] {Issue Name}
- **Domain**: {category}
- **Guideline**: {rule reference}
- **Location**: {file:line or component}
- **Problem**: {what is wrong and why it matters}
- **Fix**: {specific recommendation}
- **Code Before**: `{current code}`
- **Code After**: `{fixed code}`
### Priority Roadmap
1. {High severity items first}
2. {Medium severity items}
3. {Low severity items}
## UX Guide: {Feature/Component}
### Platform: {target platform}
### Applicable Guidelines
#### {Domain}: {Guideline Name}
- **Severity**: {level}
- **Recommendation**: {what to implement}
- **Example**: `{code example}`
- **Avoid**: `{anti-pattern}`
### Implementation Checklist
- [ ] {guideline-based check}
- [ ] {guideline-based check}
1. Identify Target → Determine what UI to review or guide
2. Detect Platform → Infer platform from code/context
3. Select Guidelines → Filter applicable rules by platform and focus
4. Analyze/Advise → Review mode: find violations | Guide mode: recommend patterns
5. Classify Severity → Rate each finding as High/Medium/Low
6. Generate Output → Produce structured report with code examples