Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when you need to find the 2-3 most popular and well-maintained npm packages relevant to a frontend checklist rule, validate they meet quality thresholds, and add them to the rule's frontmatter.
.claude/skills/thedaviddias-find-npm-packages/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -32% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -38% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -43% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -42% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -22% | 0% |
Identify, validate, and add npm package references to a Front-End Checklist rule.
Given a rule file or rule slug, this skill will:
npmPackages frontmatter fieldBefore selecting any package, check it against these criteria using the npm registry:
| Criterion | Minimum | |-----------|---------| | Weekly downloads | >10,000/week | | Last publish | Within 18 months | | npm registry | Must be listed (registry.npmjs.org/{name}) |
Packages that fail any threshold must be excluded, regardless of how relevant they are.
Read the rule file: packages/content/rules/en/{category}/{slug}.mdx
Focus on:
tools: field?Search for packages that:
Use https://www.npmjs.com/search?q={keywords} and knowledge of the ecosystem to find candidates.
Prefer packages that are:
eslint, stylelint, official plugins)For each candidate package, fetch and verify:
bash# Metadata curl https://registry.npmjs.org/{package}/latest | jq '{name, version, description}' # Weekly downloads curl https://api.npmjs.org/downloads/point/last-week/{package} | jq '.downloads'
Record the weekly download count and last publish date. Reject any below threshold.
Add a npmPackages array to the frontmatter with the package names only (metadata is fetched at build time):
yamlnpmPackages: - eslint - eslint-plugin-jsx-a11y - axe-core
Place it after tools: and before resources:.
Run the validation script to confirm all packages pass thresholds:
bashpnpm validate:packages packages/content/rules/en/{category}/{slug}.mdx
If any package fails, either remove it or replace it with a better alternative.
For the rule accessibility/form-labels:
yamlnpmPackages: - eslint-plugin-jsx-a11y - axe-core - @testing-library/jest-dom
All three have >100k weekly downloads and were published within the last 6 months.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 4,859 | 4,724 | -3% | 1 | 1 | 0% | 356 | 1,315 | +269% | 0 | 0 | — |
case-02 | fail→fail | 3,738 | 4,338 | +16% | 1 | 1 | 0% | 293 | 1,163 | +297% | 0 | 0 | — |
case-03 | fail→fail | 7,004 | 5,695 | -19% | 1 | 1 | 0% | 1,292 | 1,270 | -2% | 0 | 0 | — |
case-04 | fail→pass | 8,913 | 2,912 | -67% | 1 | 1 | 0% | 1,737 | 1,185 | -32% | 0 | 0 | — |
case-05 | fail→pass | 11,265 | 1,604 | -86% | 1 | 1 | 0% | 1,950 | 1,206 | -38% | 0 | 0 | — |
case-06 | pass→pass | 9,864 | 1,991 | -80% | 1 | 1 | 0% | 1,661 | 1,260 | -24% | 0 | 0 | — |
case-07 | pass→pass | 5,435 | 1,390 | -74% | 1 | 1 | 0% | 1,160 | 1,176 | +1% | 0 | 0 | — |
case-08 | fail→pass | 11,086 | 1,149 | -90% | 1 | 1 | 0% | 1,883 | 1,077 | -43% | 0 | 0 | — |
case-09 | pass→pass | 3,074 | 1,906 | -38% | 1 | 1 | 0% | 619 | 1,266 | +105% | 0 | 0 | — |
case-10 | pass→pass | 3,442 | 1,740 | -49% | 1 | 1 | 0% | 834 | 1,239 | +49% | 0 | 0 | — |
case-11 | fail→pass | 11,212 | 1,445 | -87% | 1 | 1 | 0% | 2,051 | 1,185 | -42% | 0 | 0 | — |
case-12 | pass→pass | 10,400 | 2,569 | -75% | 1 | 1 | 0% | 1,728 | 1,488 | -14% | 0 | 0 | — |
case-13 | pass→pass | 4,505 | 2,964 | -34% | 1 | 1 | 0% | 917 | 1,421 | +55% | 0 | 0 | — |
case-14 | fail→pass | 10,070 | 2,231 | -78% | 1 | 1 | 0% | 1,750 | 1,373 | -22% | 0 | 0 | — |
case-15 | pass→pass | 5,131 | 4,641 | -10% | 1 | 1 | 0% | 892 | 1,470 | +65% | 0 | 0 | — |
case-16 | fail→pass | 13,123 | 7,445 | -43% | 1 | 1 | 0% | 2,189 | 2,224 | +2% | 0 | 0 | — |
case-17 | pass→pass | 10,004 | 3,284 | -67% | 1 | 1 | 0% | 1,684 | 1,678 | -0% | 0 | 0 | — |
case-18 | pass→pass | 7,930 | 3,740 | -53% | 1 | 1 | 0% | 1,401 | 1,587 | +13% | 0 | 0 | — |
case-19 | fail→pass | 4,249 | 2,124 | -50% | 1 | 1 | 0% | 723 | 1,329 | +84% | 0 | 0 | — |
case-20 | pass→pass | 13,423 | 10,523 | -22% | 1 | 1 | 0% | 2,804 | 3,129 | +12% | 0 | 0 | — |
case-21 | pass→pass | 8,792 | 7,449 | -15% | 1 | 1 | 0% | 2,162 | 2,319 | +7% | 0 | 0 | — |
case-22 | pass→pass | 7,241 | 3,708 | -49% | 1 | 1 | 0% | 1,334 | 1,714 | +28% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 19 counted toward the lift figure. The other 3 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +32 percentage points is the difference between those two pass rates over the 19 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.