Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →An annotated / redline lo-fi wireframe — a desktop landing/marketing page drawn as flat greyboxes inside a browser chrome frame, overlaid with numbered annotation pins (①②③④⑤) in a single accent color, paired with a right-hand SPEC PANEL that lists each numbered pin with a short engineering / UX note. This is the "wireframe + redline spec" style — clean, flat, low-fidelity, NOT hand-drawn. Use when the brief asks for "annotated wireframe", "redline wireframe", "wireframe with spec", "lo-fi landi
.claude/skills/nexu-io-wireframe-annotated/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | 9% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 9% | 0% |
| case-20 | ✗→✓ | ▲ Improved | 9% | 0% |
| case-16 | ✓→✗ | ▼ Worse | 9% | 0% |
| case-04 | ✓→✗ | ▼ Worse | 9% | 0% |
Produce a single flat, low-fidelity landing-page wireframe with a redline spec. The whole point is "structure + handoff notes, not pixels" — greyboxes carry the layout, numbered pins call out regions, and a right-hand spec panel turns each pin into a short engineering / UX note. Keep it clean and flat; never hand-drawn or scribbly.
look. Honor type tokens only loosely (one clean sans like Inter / system-ui for labels, one mono like IBM Plex Mono for the spec notes and pin numbers). Use medium-grey fills with defined darker borders so the page reads as a thumbnail — avoid near-white-on-white, which renders blank.
framed greybox page. RIGHT is a narrow "ANNOTATIONS / SPEC" panel. Pick ONE accent color (coral or blue) and use it ONLY for the numbered pins and the matching spec numbers — everything else stays greyscale.
pin absolutely positioned on it:
left; a large image placeholder (rect with a diagonal X) on the right. Pin ②.
the accent color + a short mono/sans note, e.g. "① Sticky nav, 64px, condenses on scroll", "② Hero H1 48/56, CTA pair primary+ghost", "③ 5 partner logos, greyscale", "④ 3-up at ≥960px → 1-up mobile", "⑤ 4-col footer, legal row". Mark the panel and each row with data-od-id.
background, darken the fills/borders.
Emit between <artifact> tags:
<artifact identifier="wireframe-slug" type="text/html" title="Annotated Wireframe — Title">
<!doctype html>
<html>...</html>
</artifact>One sentence before the artifact, nothing after.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-16 | pass→fail | 39,719 | 37,747 | -5% | 1 | 1 | 0% | 8,226 | 8,932 | +9% | 0 | 0 | — |
case-01 | fail→fail | 52,991 | 113,487 | +114% | 1 | 1 | 0% | 8,265 | 8,971 | +9% | 0 | 0 | — |
case-02 | fail→fail | 54,656 | 52,414 | -4% | 1 | 1 | 0% | 8,271 | 8,977 | +9% | 0 | 0 | — |
case-03 | fail→fail | 38,322 | 49,013 | +28% | 1 | 1 | 0% | 8,267 | 8,973 | +9% | 0 | 0 | — |
case-04 | pass→fail | 49,464 | 44,090 | -11% | 1 | 1 | 0% | 8,242 | 8,948 | +9% | 0 | 0 | — |
case-05 | fail→fail | 37,910 | 52,942 | +40% | 1 | 1 | 0% | 8,230 | 8,936 | +9% | 0 | 0 | — |
case-06 | pass→fail | 28,661 | 53,775 | +88% | 1 | 1 | 0% | 4,753 | 8,936 | +88% | 0 | 0 | — |
case-07 | fail→pass | 56,418 | 36,801 | -35% | 1 | 1 | 0% | 8,234 | 8,940 | +9% | 0 | 0 | — |
case-08 | fail→fail | 35,870 | 41,286 | +15% | 1 | 1 | 0% | 8,233 | 8,939 | +9% | 0 | 0 | — |
case-09 | fail→pass | 50,340 | 35,831 | -29% | 1 | 1 | 0% | 8,225 | 8,931 | +9% | 0 | 0 | — |
case-10 | fail→fail | 50,917 | 39,534 | -22% | 1 | 1 | 0% | 8,228 | 8,934 | +9% | 0 | 0 | — |
case-11 | fail→fail | 266,933 | 41,624 | -84% | 1 | 1 | 0% | 8,223 | 8,929 | +9% | 0 | 0 | — |
case-12 | fail→fail | 51,587 | 43,963 | -15% | 1 | 1 | 0% | 8,227 | 8,933 | +9% | 0 | 0 | — |
case-13 | fail→fail | 35,978 | 48,681 | +35% | 1 | 1 | 0% | 7,674 | 8,935 | +16% | 0 | 0 | — |
case-14 | fail→fail | 39,432 | 48,034 | +22% | 1 | 1 | 0% | 8,217 | 8,923 | +9% | 0 | 0 | — |
case-15 | fail→fail | 71,197 | 47,248 | -34% | 1 | 1 | 0% | 8,225 | 8,931 | +9% | 0 | 0 | — |
case-17 | fail→fail | 37,786 | 38,266 | +1% | 1 | 1 | 0% | 8,226 | 8,932 | +9% | 0 | 0 | — |
case-18 | fail→fail | 40,550 | 38,512 | -5% | 1 | 1 | 0% | 8,219 | 8,925 | +9% | 0 | 0 | — |
case-19 | fail→fail | 36,073 | 50,024 | +39% | 1 | 1 | 0% | 8,236 | 8,942 | +9% | 0 | 0 | — |
case-20 | fail→pass | 52,776 | 49,044 | -7% | 1 | 1 | 0% | 8,229 | 8,935 | +9% | 0 | 0 | — |
case-21 | fail→fail | 37,932 | 34,580 | -9% | 1 | 1 | 0% | 8,228 | 8,934 | +9% | 0 | 0 | — |
case-22 | fail→fail | 37,728 | 49,989 | +32% | 1 | 1 | 0% | 8,218 | 8,924 | +9% | 0 | 0 | — |
case-23 | fail→fail | 52,484 | 50,231 | -4% | 1 | 1 | 0% | 8,227 | 8,933 | +9% | 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. The headline lift of 0 percentage points is the difference between those two pass rates over the 23 comparable cases. 4 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.