Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →You are a code education expert specializing in explaining complex code through clear narratives, visual diagrams, and step-by-step breakdowns. Transform difficult concepts into understandable explanations.
.claude/skills/dokhacgiakhoa-code-documentation-code-explain/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 125% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 27% | 0% |
| case-19 | ✗→✓ | ▲ Improved | -18% | 0% |
| case-18 | ✓→✗ | ▼ Worse | 3% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | 28% | 0% |
You are a code education expert specializing in explaining complex code through clear narratives, visual diagrams, and step-by-step breakdowns. Transform difficult concepts into understandable explanations for developers at all levels.
The user needs help understanding complex code sections, algorithms, design patterns, or system architectures. Focus on clarity, visual aids, and progressive disclosure of complexity to facilitate learning and onboarding.
$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 | 21,433 | 25,590 | +19% | 1 | 1 | 0% | 3,946 | 5,052 | +28% | 0 | 0 | — |
case-02 | pass→pass | 23,189 | 25,199 | +9% | 1 | 1 | 0% | 3,439 | 5,175 | +50% | 0 | 0 | — |
case-03 | fail→pass | 27,840 | 26,226 | -6% | 1 | 1 | 0% | 2,505 | 5,632 | +125% | 0 | 0 | — |
case-04 | fail→fail | 17,484 | 18,912 | +8% | 1 | 1 | 0% | 3,155 | 4,040 | +28% | 0 | 0 | — |
case-05 | fail→fail | 22,495 | 27,626 | +23% | 1 | 1 | 0% | 4,252 | 6,184 | +45% | 0 | 0 | — |
case-06 | fail→fail | 32,459 | 28,320 | -13% | 1 | 1 | 0% | 5,188 | 4,954 | -5% | 0 | 0 | — |
case-07 | pass→pass | 24,416 | 22,231 | -9% | 1 | 1 | 0% | 3,777 | 3,697 | -2% | 0 | 0 | — |
case-08 | fail→pass | 18,823 | 20,353 | +8% | 1 | 1 | 0% | 3,315 | 4,223 | +27% | 0 | 0 | — |
case-09 | pass→pass | 21,889 | 24,852 | +14% | 1 | 1 | 0% | 4,345 | 5,137 | +18% | 0 | 0 | — |
case-10 | pass→pass | 20,310 | 19,634 | -3% | 1 | 1 | 0% | 3,505 | 3,973 | +13% | 0 | 0 | — |
case-11 | fail→fail | 24,060 | 22,153 | -8% | 1 | 1 | 0% | 4,423 | 4,067 | -8% | 0 | 0 | — |
case-12 | fail→fail | 19,462 | 19,695 | +1% | 1 | 1 | 0% | 3,661 | 3,886 | +6% | 0 | 0 | — |
case-13 | pass→pass | 22,305 | 21,875 | -2% | 1 | 1 | 0% | 3,770 | 4,147 | +10% | 0 | 0 | — |
case-14 | pass→pass | 19,943 | 21,273 | +7% | 1 | 1 | 0% | 3,575 | 3,903 | +9% | 0 | 0 | — |
case-15 | pass→pass | 31,529 | 21,442 | -32% | 1 | 1 | 0% | 4,801 | 4,530 | -6% | 0 | 0 | — |
case-16 | fail→fail | 20,505 | 23,524 | +15% | 1 | 1 | 0% | 3,592 | 5,011 | +40% | 0 | 0 | — |
case-17 | fail→fail | 20,719 | 18,480 | -11% | 1 | 1 | 0% | 3,043 | 3,712 | +22% | 0 | 0 | — |
case-18 | pass→fail | 23,118 | 17,516 | -24% | 1 | 1 | 0% | 3,515 | 3,636 | +3% | 0 | 0 | — |
case-19 | fail→pass | 23,144 | 21,013 | -9% | 1 | 1 | 0% | 4,579 | 3,767 | -18% | 0 | 0 | — |
case-20 | pass→pass | 21,245 | 23,485 | +11% | 1 | 1 | 0% | 4,192 | 4,024 | -4% | 0 | 0 | — |
case-21 | pass→pass | 12,369 | 21,842 | +77% | 1 | 1 | 0% | 3,006 | 4,211 | +40% | 0 | 0 | — |
case-22 | pass→pass | 18,363 | 20,078 | +9% | 1 | 1 | 0% | 3,617 | 3,771 | +4% | 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 +9 percentage points is the difference between those two pass rates over the 22 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.