Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Write comprehensive technical documentation including user guides, how-to articles, system architecture docs, onboarding materials, and knowledge base articles. Creates clear, structured documentation for technical and non-technical audiences. Use when users need technical writing, documentation, tutorials, or knowledge base content.
.claude/skills/onewave-ai-technical-writer/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -33% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 26% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -16% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 32% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 20% | 0% |
Create clear, comprehensive technical documentation for any audience.
references/templates.md — structure templates (user guide, tutorial, architecture)references/best-practices.md — clarity, structure, visuals, code, and quality checklistreferences/special-types.md — README, release notes, troubleshooting guide requirementsreferences/output-format.md — final delivery wrapper and maintenance notesreferences/special-types.md for the full type list and type-specific section requirements.references/templates.md. Start simple, then layer in complexity.references/best-practices.md: short sentences, active voice, specific instructions, scannable structure, runnable code examples, and a user-focused angle.references/best-practices.md.references/output-format.md, including the documentation checklist and maintenance notes.Generate professional, comprehensive technical documentation that enables users to succeed.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 29,373 | 18,188 | -38% | 1 | 1 | 0% | 5,508 | 3,610 | -34% | 0 | 0 | — |
case-02 | pass→fail | 22,261 | 63,534 | +185% | 1 | 1 | 0% | 3,758 | 1,345 | -64% | 0 | 0 | — |
case-03 | pass→pass | 42,435 | 25,441 | -40% | 1 | 1 | 0% | 7,581 | 5,111 | -33% | 0 | 0 | — |
case-04 | fail→pass | 41,907 | 26,627 | -36% | 1 | 1 | 0% | 8,231 | 5,485 | -33% | 0 | 0 | — |
case-05 | pass→pass | 19,707 | 24,381 | +24% | 1 | 1 | 0% | 3,781 | 4,895 | +29% | 0 | 0 | — |
case-06 | fail→pass | 15,150 | 18,478 | +22% | 1 | 1 | 0% | 3,004 | 3,788 | +26% | 0 | 0 | — |
case-07 | fail→fail | 30,024 | 22,434 | -25% | 1 | 1 | 0% | 5,260 | 4,240 | -19% | 0 | 0 | — |
case-08 | pass→pass | 23,826 | 27,773 | +17% | 1 | 1 | 0% | 3,804 | 4,867 | +28% | 0 | 0 | — |
case-09 | pass→pass | 18,538 | 23,497 | +27% | 1 | 1 | 0% | 3,505 | 4,729 | +35% | 0 | 0 | — |
case-10 | fail→pass | 33,479 | 26,054 | -22% | 1 | 1 | 0% | 6,499 | 5,485 | -16% | 0 | 0 | — |
case-11 | fail→pass | 25,915 | 29,100 | +12% | 1 | 1 | 0% | 4,666 | 6,170 | +32% | 0 | 0 | — |
case-12 | fail→fail | 23,243 | 25,150 | +8% | 1 | 1 | 0% | 4,248 | 4,796 | +13% | 0 | 0 | — |
case-13 | fail→pass | 28,435 | 30,893 | +9% | 1 | 1 | 0% | 4,802 | 5,782 | +20% | 0 | 0 | — |
case-14 | fail→pass | 22,569 | 23,642 | +5% | 1 | 1 | 0% | 3,864 | 4,644 | +20% | 0 | 0 | — |
case-15 | fail→pass | 28,068 | 32,552 | +16% | 1 | 1 | 0% | 5,191 | 6,387 | +23% | 0 | 0 | — |
case-16 | fail→fail | 17,766 | 21,037 | +18% | 1 | 1 | 0% | 3,406 | 4,242 | +25% | 0 | 0 | — |
case-17 | fail→pass | 23,260 | 26,765 | +15% | 1 | 1 | 0% | 4,030 | 5,654 | +40% | 0 | 0 | — |
case-18 | fail→pass | 15,332 | 19,212 | +25% | 1 | 1 | 0% | 2,587 | 3,730 | +44% | 0 | 0 | — |
case-19 | pass→pass | 25,160 | 29,217 | +16% | 1 | 1 | 0% | 4,492 | 5,676 | +26% | 0 | 0 | — |
case-20 | fail→fail | 4,801 | 12,926 | +169% | 1 | 1 | 0% | 937 | 2,666 | +185% | 0 | 0 | — |
case-21 | fail→fail | 9,725 | 12,913 | +33% | 1 | 1 | 0% | 1,781 | 2,734 | +54% | 0 | 0 | — |
case-22 | pass→fail | 10,717 | 18,021 | +68% | 1 | 1 | 0% | 2,072 | 3,698 | +78% | 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 21 counted toward the lift figure. The other 1 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 21 comparable cases. 2 cases got worse with the skill loaded, and they are 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.