ux-evaluator — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited ux-evaluator (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Systematically evaluate UI components against established UX principles using a 3-dimension framework. Transform subjective design feedback into actionable, evidence-based recommendations by comparing against industry conventions and authoritative sources.
Core Value: Prevents subjective design debates by grounding decisions in documented best practices.
Triggers:
Do NOT use for:
For ANY UI component, evaluate these three dimensions:
| Dimension | What to Analyze | Key Questions |
|---|---|---|
| 1. Position | Where is it located relative to other elements? | Does position follow conventions? Is it discoverable? |
| 2. Visual Weight | How prominent is it visually? | Does it compete with primary actions? Is hierarchy clear? |
| 3. Spacing | What's the gap from adjacent elements? | Is there adequate separation? Is spacing consistent? |
Step 1: GATHER CONTEXT
├── What component is being evaluated?
├── What user feedback or concern triggered this?
├── Is there an external reference (article, guideline)?
└── What is the component's purpose (primary CTA, utility, navigation)?
Step 2: ANALYZE CURRENT STATE
├── Position: Document exact location in layout
├── Visual Weight: Describe styling (filled, ghost, icon-only, etc.)
├── Spacing: Measure gaps from adjacent elements
└── Compare to industry conventions (see references/)
Step 3: PRODUCE VERDICT
├── For each dimension: CORRECT / NEEDS CHANGE / ACCEPTABLE
├── If NEEDS CHANGE: Specific recommendation with rationale
├── Reference authoritative source for each recommendation
└── Prioritize changes (P1: breaks UX, P2: suboptimal, P3: polish)Position:
Visual Weight:
Spacing:
Labels:
Position:
Visual Weight:
Spacing:
Position:
Visual Weight:
Spacing:
| Site | Pattern |
|---|---|
| GitHub | [Sign In] [Sign Up] - secondary left, primary right |
| Stripe | [Sign In] [Start now →] - secondary left, primary right |
| [Sign In] [Create account] - same pattern | |
| Notion | [Log in] [Get Notion free] - same pattern |
Verdict: Secondary LEFT, Primary RIGHT is the standard.
| Site | Placement |
|---|---|
| GitHub | Far right, after user menu |
| VS Code Docs | Far right |
| Stripe Docs | Far right |
| Discord | In settings, not navbar |
Verdict: Far right (after auth) or in settings dropdown.
| Control | Expected Weight |
|---|---|
| Theme toggle | Icon-only, subtle, doesn't compete with CTAs |
| Search | Icon trigger or compact input, expandable |
| Language selector | Icon or compact dropdown |
Verdict: Utilities should be accessible but subordinate to primary actions.
When evaluating a component, produce this structured output:
## [Component Name] Evaluation
### Current State
- **Position**: [Description]
- **Visual Weight**: [Description]
- **Spacing**: [Measurements]
### Analysis
| Dimension | Assessment | Rationale |
|-----------|------------|-----------|
| Position | [OK/CHANGE] | [Why, with reference] |
| Visual Weight | [OK/CHANGE] | [Why, with reference] |
| Spacing | [OK/CHANGE] | [Why, with reference] |
### Verdict: [CORRECT / NEEDS CHANGES]
### Recommendations (if any)
| Priority | Change | Rationale |
|----------|--------|-----------|
| P1 | [Specific change] | [Reference to principle] |
| P2 | [Specific change] | [Reference to principle] |See references/ for detailed UX principles:
balsamiq-button-principles.md - Button design best practicesnielsen-heuristics.md - 10 usability heuristics (to be added)Before finalizing evaluation:
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.