Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Structure compelling presentations with clear narratives and visual suggestions
.claude/skills/nicepkg-presentation-builder/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-09 | ✗→✓ | ▲ Improved | -11% | 0% |
| case-22 | ✗→✓ | ▲ Improved | 41% | 0% |
| case-21 | ✓→✗ | ▼ Worse | -1% | 0% |
| case-04 | ✓→✓ | = Same ✓ | -8% | 0% |
| case-12 | ✓→✓ | = Same ✓ | -11% | 0% |
Create presentation outlines with storytelling structure and visual guidance.
Presentation: [Title]
Audience: [Type] | Duration: [Minutes]
Key Message: [One sentence]
Slide 1: Title Slide
- Title: [Compelling title]
- Subtitle: [Context]
- Visual: [Logo/image suggestion]
Slide 2: Hook/Problem
- Headline: [Attention grabber]
- Content: [Bullet points]
- Visual: [Chart/image type]
- Speaker notes: [What to say]
Slide 3-N: [Main Content]
...
Slide N-1: Solution/Recommendation
- Headline: [Action-oriented]
- Content: [Key points]
- Visual: [Before/after, diagram]
Final Slide: Call to Action
- The Ask: [Specific request]
- Next Steps: [Clear actions]
---
Presentation Flow:
Problem → Evidence → Solution → Benefits → Ask
Timing Guide:
- Intro: 2 min
- Body: [N] min
- Q&A: [N] min| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 24,597 | 27,451 | +12% | 1 | 1 | 0% | 3,878 | 2,882 | -26% | 0 | 0 | — |
case-02 | fail→fail | 28,283 | 12,831 | -55% | 1 | 1 | 0% | 4,557 | 2,403 | -47% | 0 | 0 | — |
case-03 | fail→fail | 22,972 | 12,634 | -45% | 1 | 1 | 0% | 3,485 | 2,419 | -31% | 0 | 0 | — |
case-04 | pass→pass | 15,564 | 13,419 | -14% | 1 | 1 | 0% | 2,586 | 2,370 | -8% | 0 | 0 | — |
case-05 | fail→fail | 17,122 | 16,025 | -6% | 1 | 1 | 0% | 2,588 | 2,794 | +8% | 0 | 0 | — |
case-06 | fail→fail | 31,533 | 24,061 | -24% | 1 | 1 | 0% | 6,182 | 4,277 | -31% | 0 | 0 | — |
case-07 | fail→fail | 14,822 | 11,670 | -21% | 1 | 1 | 0% | 2,502 | 2,145 | -14% | 0 | 0 | — |
case-08 | fail→fail | 12,116 | 12,318 | +2% | 1 | 1 | 0% | 1,938 | 2,213 | +14% | 0 | 0 | — |
case-09 | fail→pass | 17,269 | 13,529 | -22% | 1 | 1 | 0% | 2,817 | 2,517 | -11% | 0 | 0 | — |
case-10 | fail→fail | 9,229 | 8,781 | -5% | 1 | 1 | 0% | 1,491 | 1,704 | +14% | 0 | 0 | — |
case-11 | fail→fail | 16,203 | 12,400 | -23% | 1 | 1 | 0% | 2,644 | 2,383 | -10% | 0 | 0 | — |
case-12 | pass→pass | 23,514 | 19,085 | -19% | 1 | 1 | 0% | 3,857 | 3,415 | -11% | 0 | 0 | — |
case-13 | fail→fail | 14,837 | 11,720 | -21% | 1 | 1 | 0% | 2,291 | 2,097 | -8% | 0 | 0 | — |
case-14 | fail→fail | 17,097 | 12,923 | -24% | 1 | 1 | 0% | 2,602 | 2,318 | -11% | 0 | 0 | — |
case-15 | fail→fail | 26,615 | 12,877 | -52% | 1 | 1 | 0% | 4,060 | 2,338 | -42% | 0 | 0 | — |
case-16 | fail→fail | 14,623 | 12,555 | -14% | 1 | 1 | 0% | 2,423 | 2,313 | -5% | 0 | 0 | — |
case-17 | fail→fail | 15,788 | 10,053 | -36% | 1 | 1 | 0% | 2,316 | 2,044 | -12% | 0 | 0 | — |
case-18 | fail→fail | 21,954 | 17,073 | -22% | 1 | 1 | 0% | 2,925 | 2,900 | -1% | 0 | 0 | — |
case-19 | pass→pass | 27,615 | 14,087 | -49% | 1 | 1 | 0% | 6,173 | 3,318 | -46% | 0 | 0 | — |
case-20 | pass→pass | 18,344 | 18,857 | +3% | 1 | 1 | 0% | 3,883 | 4,386 | +13% | 0 | 0 | — |
case-21 | pass→fail | 19,407 | 17,769 | -8% | 1 | 1 | 0% | 2,952 | 2,919 | -1% | 0 | 0 | — |
case-22 | fail→pass | 13,976 | 17,857 | +28% | 1 | 1 | 0% | 2,267 | 3,203 | +41% | 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. 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.