Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Create stunning, animation-rich HTML presentations from scratch or by converting PowerPoint files. Use when the user wants to build a presentation, convert a PPT/PPTX to web, or create slides for a talk/pitch. Helps non-designers discover their aesthetic through visual exploration rather than abstract choices.
.claude/skills/loulanyue-frontend-slides/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | 66% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 57% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -3% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 44% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 31% | 0% |
Create zero-dependency, animation-rich HTML presentations that run entirely in the browser.
Inspired by the visual exploration approach showcased in work by zarazhangrui (credit: @zarazhangrui).
.ppt or .pptx slides into an HTML presentationBefore generating, read STYLE_PRESETS.md for the viewport-safe CSS base, density limits, preset catalog, and CSS gotchas.
Choose one path:
.ppt or .pptxAsk only the minimum needed:
If the user has content, ask them to paste it before styling.
Default to visual exploration.
If the user already knows the desired preset, skip previews and use it directly.
Otherwise:
.ecc-design/slide-previews/.Use the preset guide in STYLE_PRESETS.md when mapping mood to style.
Output either:
presentation.html[presentation-name].htmlUse an assets/ folder only when the deck contains extracted or user-supplied images.
Required structure:
STYLE_PRESETS.mdTreat this as a hard gate.
Rules:
.slide must use height: 100vh; height: 100dvh; overflow: hidden;clamp()Use the density limits and mandatory CSS block in STYLE_PRESETS.md.
Check the finished deck at these sizes:
If browser automation is available, use it to verify no slide overflows and that keyboard navigation works.
At handoff:
Use the correct opener for the current OS:
open file.htmlxdg-open file.htmlstart "" file.htmlFor PowerPoint conversion:
python3 with python-pptx to extract text, images, and notes.python-pptx is unavailable, ask whether to install it or fall back to a manual/export-based workflow.Keep conversion cross-platform. Do not rely on macOS-only tools when Python can do the job.
Include:
main, section, nav)prefers-reduced-motionUse these maxima unless the user explicitly asks for denser slides and readability still holds:
| Slide type | Limit | |------------|-------| | Title | 1 heading + 1 subtitle + optional tagline | | Content | 1 heading + 4-6 bullets or 2 short paragraphs | | Feature grid | 6 cards max | | Code | 8-10 lines max | | Quote | 1 quote + attribution | | Image | 1 image constrained by viewport |
-clamp(...)frontend-patterns for component and interaction patterns around the deckliquid-glass-design when a presentation intentionally borrows Apple glass aestheticse2e-testing if you need automated browser verification for the final deck| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-02 | fail→fail | 10,531 | 9,130 | -13% | 1 | 1 | 0% | 1,776 | 3,098 | +74% | 0 | 0 | — |
case-01 | fail→fail | 8,271 | 4,006 | -52% | 1 | 1 | 0% | 1,087 | 2,009 | +85% | 0 | 0 | — |
case-03 | fail→fail | 21,257 | 23,697 | +11% | 1 | 1 | 0% | 4,005 | 6,670 | +67% | 0 | 0 | — |
case-04 | fail→fail | 20,075 | 10,177 | -49% | 1 | 1 | 0% | 4,223 | 3,230 | -24% | 0 | 0 | — |
case-05 | fail→fail | 8,678 | 4,211 | -51% | 1 | 1 | 0% | 1,458 | 2,255 | +55% | 0 | 0 | — |
case-06 | fail→fail | 12,576 | 14,129 | +12% | 1 | 1 | 0% | 1,909 | 3,956 | +107% | 0 | 0 | — |
case-07 | fail→pass | 14,274 | 11,476 | -20% | 1 | 1 | 0% | 2,175 | 3,620 | +66% | 0 | 0 | — |
case-08 | pass→pass | 13,559 | 12,973 | -4% | 1 | 1 | 0% | 2,221 | 3,989 | +80% | 0 | 0 | — |
case-09 | fail→pass | 17,144 | 15,407 | -10% | 1 | 1 | 0% | 2,840 | 4,445 | +57% | 0 | 0 | — |
case-10 | fail→pass | 15,657 | 5,796 | -63% | 1 | 1 | 0% | 2,565 | 2,487 | -3% | 0 | 0 | — |
case-11 | pass→pass | 28,039 | 28,839 | +3% | 1 | 1 | 0% | 6,193 | 7,755 | +25% | 0 | 0 | — |
case-12 | pass→pass | 14,115 | 11,925 | -16% | 1 | 1 | 0% | 2,777 | 3,797 | +37% | 0 | 0 | — |
case-13 | pass→fail | 28,910 | 27,256 | -6% | 1 | 1 | 0% | 6,185 | 7,747 | +25% | 0 | 0 | — |
case-14 | fail→pass | 13,047 | 10,200 | -22% | 1 | 1 | 0% | 2,218 | 3,185 | +44% | 0 | 0 | — |
case-15 | fail→fail | 12,709 | 10,344 | -19% | 1 | 1 | 0% | 1,949 | 3,067 | +57% | 0 | 0 | — |
case-16 | pass→pass | 12,527 | 11,739 | -6% | 1 | 1 | 0% | 2,265 | 3,671 | +62% | 0 | 0 | — |
case-17 | fail→pass | 15,259 | 10,662 | -30% | 1 | 1 | 0% | 2,670 | 3,488 | +31% | 0 | 0 | — |
case-18 | pass→pass | 17,664 | 22,898 | +30% | 1 | 1 | 0% | 3,115 | 6,254 | +101% | 0 | 0 | — |
case-19 | pass→pass | 11,137 | 9,284 | -17% | 1 | 1 | 0% | 1,821 | 3,230 | +77% | 0 | 0 | — |
case-20 | fail→fail | 11,192 | 6,612 | -41% | 1 | 1 | 0% | 1,719 | 2,642 | +54% | 0 | 0 | — |
case-21 | pass→pass | 10,795 | 4,771 | -56% | 1 | 1 | 0% | 1,836 | 2,401 | +31% | 0 | 0 | — |
case-22 | pass→pass | 16,014 | 12,029 | -25% | 1 | 1 | 0% | 2,518 | 3,621 | +44% | 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 +18 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.