Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use this plugin when the user wants a "Bold Signal Frame" HyperFrames motion video — Bold colored card on a dark gradient — big section number, nav breadcrumb, orange card sliding in, title rising.
.claude/skills/nexu-io-video-template-frame-bold-signal/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -69% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -79% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -59% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -30% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -67% | 0% |
Bold colored card on a dark gradient — big section number, nav breadcrumb, orange card sliding in, title rising.
A HyperFrames-ready HTML + CSS + GSAP motion composition, bundled under source/. It renders deterministically to MP4 / WEBM at 16:9, 9:16, 1:1, default 15s, 60fps.
Best for: Section / chapter divider · Bold launch statement · High-impact title card
source/index.html to understand the named layers and the animation timeline.source/; do not introduce external network assets that would break a headless render.Source: html-video templates/frame-bold-signal (license Apache-2.0). Derived from frontend-slides (Zara Zhang, MIT) — https://github.com/zarazhangrui/frontend-slides.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-03 | fail→fail | 34,654 | 24,632 | -29% | 1 | 1 | 0% | 6,409 | 5,471 | -15% | 0 | 0 | — |
case-01 | fail→fail | 16,776 | 12,751 | -24% | 1 | 1 | 0% | 2,957 | 465 | -84% | 0 | 0 | — |
case-02 | fail→fail | 40,314 | 59,179 | +47% | 1 | 1 | 0% | 8,239 | 13,741 | +67% | 0 | 0 | — |
case-04 | fail→pass | 10,016 | 1,894 | -81% | 1 | 1 | 0% | 1,766 | 552 | -69% | 0 | 0 | — |
case-05 | pass→pass | 8,215 | 3,207 | -61% | 1 | 1 | 0% | 1,193 | 775 | -35% | 0 | 0 | — |
case-06 | pass→pass | 13,383 | 7,844 | -41% | 1 | 1 | 0% | 2,134 | 1,562 | -27% | 0 | 0 | — |
case-07 | fail→pass | 14,771 | 1,430 | -90% | 1 | 1 | 0% | 2,216 | 469 | -79% | 0 | 0 | — |
case-08 | fail→pass | 7,948 | 1,637 | -79% | 1 | 1 | 0% | 1,233 | 506 | -59% | 0 | 0 | — |
case-09 | pass→pass | 12,364 | 5,641 | -54% | 1 | 1 | 0% | 1,964 | 1,121 | -43% | 0 | 0 | — |
case-10 | pass→pass | 8,618 | 6,400 | -26% | 1 | 1 | 0% | 1,536 | 1,424 | -7% | 0 | 0 | — |
case-11 | pass→pass | 11,455 | 13,421 | +17% | 1 | 1 | 0% | 2,264 | 2,689 | +19% | 0 | 0 | — |
case-12 | pass→pass | 21,509 | 34,071 | +58% | 1 | 1 | 0% | 4,542 | 7,475 | +65% | 0 | 0 | — |
case-13 | fail→pass | 10,574 | 5,767 | -45% | 1 | 1 | 0% | 1,586 | 1,117 | -30% | 0 | 0 | — |
case-14 | fail→pass | 9,287 | 1,638 | -82% | 1 | 1 | 0% | 1,434 | 479 | -67% | 0 | 0 | — |
case-15 | pass→pass | 11,590 | 10,426 | -10% | 1 | 1 | 0% | 1,735 | 2,038 | +17% | 0 | 0 | — |
case-16 | pass→pass | 10,570 | 1,567 | -85% | 1 | 1 | 0% | 1,488 | 466 | -69% | 0 | 0 | — |
case-17 | fail→pass | 10,360 | 4,456 | -57% | 1 | 1 | 0% | 1,466 | 971 | -34% | 0 | 0 | — |
case-18 | fail→pass | 6,594 | 1,877 | -72% | 1 | 1 | 0% | 954 | 532 | -44% | 0 | 0 | — |
case-19 | fail→pass | 38,289 | 1,625 | -96% | 1 | 1 | 0% | 6,482 | 455 | -93% | 0 | 0 | — |
case-20 | pass→pass | 4,967 | 1,513 | -70% | 1 | 1 | 0% | 724 | 459 | -37% | 0 | 0 | — |
case-21 | fail→pass | 10,702 | 3,478 | -68% | 1 | 1 | 0% | 1,611 | 784 | -51% | 0 | 0 | — |
case-22 | pass→pass | 12,313 | 9,612 | -22% | 1 | 1 | 0% | 1,831 | 1,997 | +9% | 0 | 0 | — |
case-23 | fail→pass | 10,540 | 4,556 | -57% | 1 | 1 | 0% | 1,741 | 1,105 | -37% | 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. 23 cases were attempted, and 21 counted toward the lift figure. The other 2 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 +43 percentage points is the difference between those two pass rates over the 21 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.