Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →You are a documentation expert specializing in creating comprehensive, maintainable documentation from code. Generate API docs, architecture diagrams, user guides, and technical references using AI-powered analysis and industry best practices.
.claude/skills/dokhacgiakhoa-documentation-generation-doc-generate/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 10% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | 8% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 24% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | 25% | 0% |
| case-06 | ✓→✓ | ＝ Same ✓ | 18% | 0% |
You are a documentation expert specializing in creating comprehensive, maintainable documentation from code. Generate API docs, architecture diagrams, user guides, and technical references using AI-powered analysis and industry best practices.
The user needs automated documentation generation that extracts information from code, creates clear explanations, and maintains consistency across documentation types. Focus on creating living documentation that stays synchronized with code.
$ARGUMENTS
resources/implementation-playbook.md.resources/implementation-playbook.md for detailed examples and templates.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 8,532 | 19,081 | +124% | 1 | 1 | 0% | 1,479 | 1,596 | +8% | 0 | 0 | — |
case-02 | fail→fail | 8,842 | 12,185 | +38% | 1 | 1 | 0% | 1,470 | 2,562 | +74% | 0 | 0 | — |
case-03 | fail→pass | 19,853 | 18,717 | -6% | 1 | 1 | 0% | 3,432 | 3,792 | +10% | 0 | 0 | — |
case-04 | pass→pass | 11,563 | 11,462 | -1% | 1 | 1 | 0% | 1,977 | 2,458 | +24% | 0 | 0 | — |
case-05 | pass→pass | 15,907 | 16,146 | +2% | 1 | 1 | 0% | 2,799 | 3,499 | +25% | 0 | 0 | — |
case-06 | pass→pass | 16,938 | 20,431 | +21% | 1 | 1 | 0% | 2,871 | 3,375 | +18% | 0 | 0 | — |
case-07 | pass→pass | 18,014 | 16,988 | -6% | 1 | 1 | 0% | 2,501 | 3,405 | +36% | 0 | 0 | — |
case-08 | pass→pass | 17,485 | 20,329 | +16% | 1 | 1 | 0% | 2,546 | 3,452 | +36% | 0 | 0 | — |
case-09 | pass→pass | 15,365 | 18,460 | +20% | 1 | 1 | 0% | 2,841 | 3,794 | +34% | 0 | 0 | — |
case-10 | pass→pass | 18,625 | 22,900 | +23% | 1 | 1 | 0% | 2,542 | 4,693 | +85% | 0 | 0 | — |
case-11 | pass→pass | 11,306 | 14,833 | +31% | 1 | 1 | 0% | 1,814 | 3,263 | +80% | 0 | 0 | — |
case-12 | pass→pass | 14,548 | 30,514 | +110% | 1 | 1 | 0% | 2,574 | 5,240 | +104% | 0 | 0 | — |
case-13 | pass→pass | 15,380 | 17,727 | +15% | 1 | 1 | 0% | 2,603 | 3,625 | +39% | 0 | 0 | — |
case-14 | pass→pass | 20,775 | 15,652 | -25% | 1 | 1 | 0% | 2,869 | 2,882 | +0% | 0 | 0 | — |
case-15 | pass→pass | 16,296 | 18,703 | +15% | 1 | 1 | 0% | 3,307 | 3,859 | +17% | 0 | 0 | — |
case-16 | pass→pass | 12,588 | 17,825 | +42% | 1 | 1 | 0% | 2,092 | 2,945 | +41% | 0 | 0 | — |
case-17 | pass→pass | 21,121 | 17,524 | -17% | 1 | 1 | 0% | 3,302 | 3,073 | -7% | 0 | 0 | — |
case-18 | pass→pass | 18,853 | 19,769 | +5% | 1 | 1 | 0% | 2,961 | 3,552 | +20% | 0 | 0 | — |
case-19 | pass→pass | 11,464 | 21,434 | +87% | 1 | 1 | 0% | 1,906 | 3,550 | +86% | 0 | 0 | — |
case-20 | pass→pass | 19,227 | 18,932 | -2% | 1 | 1 | 0% | 2,627 | 3,600 | +37% | 0 | 0 | — |
case-21 | pass→pass | 19,297 | 20,937 | +8% | 1 | 1 | 0% | 2,965 | 3,502 | +18% | 0 | 0 | — |
case-22 | pass→pass | 12,286 | 16,498 | +34% | 1 | 1 | 0% | 2,215 | 2,757 | +24% | 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. The headline lift of +5 percentage points is the difference between those two pass rates over the 22 comparable cases.
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.