Quality Score Methodology
Every skill in the FindUtils directory is scored across 5 dimensions on a 1-10 scale. The scores produce a letter rank (S/A/B/C) and determine whether a skill is Featured, Recommended, or Community tier. Here is exactly how it works.
The 5 Dimensions
Practicality
Does this skill solve a real, common problem? Will a developer actually use it in production?
1 = theoretical • 10 = solves daily pain
Clarity
Is the rule clearly written? Can an AI assistant follow it without ambiguity?
1 = vague • 10 = unambiguous instructions
Actionability
Does the rule produce measurably better AI output? Specific code patterns, not just philosophy.
1 = generic advice • 10 = concrete patterns
Freshness
Is the content verified against current framework versions? Does it reference latest APIs?
1 = outdated • 10 = verified this month
Impact
How much does this skill improve AI output quality? High-impact rules transform entire workflows.
1 = marginal • 10 = transformative
How Ranks Are Computed
The 5 dimension scores are averaged to produce a composite score. The rank is assigned based on the composite:
| Rank | Composite Score | Meaning |
|---|---|---|
| S | 9.0 - 10.0 | Exceptional. Best-in-class rules that every developer should use. |
| A | 7.0 - 8.9 | Excellent. Production-ready rules with strong coverage. |
| B | 5.0 - 6.9 | Good. Useful rules that cover the basics well. |
| C | Below 5.0 | Needs improvement. May be outdated or too generic. |
Quality Tiers
★ Featured
Rank S or A with a Freshness score of 8+. Manually reviewed and verified against the latest framework versions. These are the skills we recommend first.
✓ Recommended
Rank A or B with solid scores across all dimensions. Reliable and useful, though they may not cover every edge case.
Community
User-submitted skills that have passed basic quality checks but haven't been fully reviewed. Quality varies.
Freshness Verification
Skills that target specific frameworks (React, Next.js, Tailwind, etc.) are periodically verified against the latest stable versions. The verification records:
- Date verified -- when the skill was last tested
- Versions tested -- the specific framework versions (e.g., React 19.2.0, TypeScript 5.6.0)
A skill verified within the last 60 days shows a green "Fresh" badge. 60-180 days shows yellow. Over 180 days shows red "Stale."
How to Improve a Skill's Score
If you want to contribute or improve a skill:
- Be specific. "Use React hooks" scores low on Actionability. "Use useMemo for expensive computations, useCallback for event handlers passed to memoized children" scores high.
- Reference versions. "Works with React" scores low on Freshness. "Verified against React 19.2.0 with Server Components" scores high.
- Provide patterns, not philosophy. AI assistants need concrete code patterns. Include function signatures, file structures, naming conventions.
- Cover edge cases. Rules that handle error states, empty inputs, and loading states score higher on Practicality.
- Stay current. Update your rules when major framework versions ship. Outdated rules lose Freshness points fast.