Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Style guidelines for writing and updating documentation. Use when writing new docs, updating existing docs, or reviewing docs for quality.
.claude/skills/bilal140202-docs-style/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 27% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 319% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 66% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 20% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 1% | 0% |
<Tip>, <Info>, <Warning>, etc. These components break reading flow and look heavy when overused. Reserve them for truly critical warnings (e.g. data loss, security). One per page is a good ceiling; zero is often fine.> Tell me about the CLI but > @tests/auth.test.ts This test started failing after the last migration.title in frontmatter should match the sidebar label. Drop sidebarTitle unless there's a genuine reason for them to differ. Don't stuff extra context into the page title (e.g., "Continue CLI (cn) Overview" → "Overview").description field. The opening paragraph of the page should provide whatever context is needed. Metadata subtitles add clutter and duplicate what the prose already says.@ Context"). The test isn't grammatical — it's tonal. If it sounds like a friendly narrator is walking you through something, rewrite it. If it just plainly states what the section covers, it's good.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-04 | pass→fail | 13,212 | 12,471 | -6% | 1 | 1 | 0% | 2,043 | 3,297 | +61% | 0 | 0 | — |
case-01 | fail→pass | 15,160 | 12,051 | -21% | 1 | 1 | 0% | 2,631 | 3,343 | +27% | 0 | 0 | — |
case-02 | fail→fail | 17,141 | 13,569 | -21% | 1 | 1 | 0% | 3,264 | 3,848 | +18% | 0 | 0 | — |
case-03 | fail→fail | 14,125 | 13,797 | -2% | 1 | 1 | 0% | 2,906 | 3,837 | +32% | 0 | 0 | — |
case-05 | pass→pass | 15,006 | 14,982 | -0% | 1 | 1 | 0% | 3,643 | 4,437 | +22% | 0 | 0 | — |
case-06 | pass→pass | 15,923 | 12,520 | -21% | 1 | 1 | 0% | 2,737 | 3,168 | +16% | 0 | 0 | — |
case-07 | fail→pass | 5,499 | 16,347 | +197% | 1 | 1 | 0% | 989 | 4,144 | +319% | 0 | 0 | — |
case-08 | fail→pass | 12,162 | 12,952 | +6% | 1 | 1 | 0% | 2,149 | 3,578 | +66% | 0 | 0 | — |
case-09 | fail→pass | 16,108 | 13,152 | -18% | 1 | 1 | 0% | 3,087 | 3,710 | +20% | 0 | 0 | — |
case-10 | fail→fail | 7,312 | 19,244 | +163% | 1 | 1 | 0% | 1,427 | 4,886 | +242% | 0 | 0 | — |
case-11 | fail→fail | 19,142 | 16,335 | -15% | 1 | 1 | 0% | 3,681 | 4,397 | +19% | 0 | 0 | — |
case-12 | fail→pass | 17,404 | 11,352 | -35% | 1 | 1 | 0% | 3,076 | 3,114 | +1% | 0 | 0 | — |
case-13 | fail→pass | 4,231 | 4,934 | +17% | 1 | 1 | 0% | 749 | 1,907 | +155% | 0 | 0 | — |
case-14 | pass→pass | 21,647 | 16,483 | -24% | 1 | 1 | 0% | 4,385 | 4,524 | +3% | 0 | 0 | — |
case-15 | fail→pass | 9,637 | 7,295 | -24% | 1 | 1 | 0% | 1,768 | 2,373 | +34% | 0 | 0 | — |
case-16 | fail→fail | 12,863 | 12,707 | -1% | 1 | 1 | 0% | 2,489 | 3,531 | +42% | 0 | 0 | — |
case-17 | fail→pass | 13,631 | 12,352 | -9% | 1 | 1 | 0% | 2,274 | 3,229 | +42% | 0 | 0 | — |
case-18 | fail→fail | 17,800 | 10,768 | -40% | 1 | 1 | 0% | 3,317 | 2,961 | -11% | 0 | 0 | — |
case-19 | fail→fail | 6,501 | 11,163 | +72% | 1 | 1 | 0% | 1,144 | 2,911 | +154% | 0 | 0 | — |
case-20 | fail→pass | 10,976 | 15,687 | +43% | 1 | 1 | 0% | 2,154 | 4,124 | +91% | 0 | 0 | — |
case-21 | fail→fail | 16,676 | 12,019 | -28% | 1 | 1 | 0% | 3,024 | 3,292 | +9% | 0 | 0 | — |
case-22 | fail→pass | 19,944 | 16,724 | -16% | 1 | 1 | 0% | 4,216 | 4,405 | +4% | 0 | 0 | — |
case-23 | fail→fail | 8,082 | 13,644 | +69% | 1 | 1 | 0% | 1,629 | 3,817 | +134% | 0 | 0 | — |
case-24 | fail→pass | 8,416 | 9,194 | +9% | 1 | 1 | 0% | 1,534 | 2,721 | +77% | 0 | 0 | — |
case-25 | fail→pass | 12,381 | 15,450 | +25% | 1 | 1 | 0% | 2,601 | 4,219 | +62% | 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. 25 cases were attempted. The headline lift of +44 percentage points is the difference between those two pass rates over the 25 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.