Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A consumer-feeling dating / matchmaking dashboard — left rail navigation, ticker bar of community signals, headline KPIs, a 30-day mutual-matches bar chart, and a match-rate trend block. Editorial typography, restrained accent. Use when the brief asks for a "dating site", "matchmaking", "community dashboard", "social network dashboard", or any consumer product where the data is the story.
.claude/skills/nexu-io-dating-web/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-20 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-05 | ✓→✗ | ▼ Worse | 13% | 0% |
| case-04 | ✓→✗ | ▼ Worse | 13% | 0% |
| case-06 | ✓→✗ | ▼ Worse | 129% | 0% |
Produce a single-screen consumer dashboard that feels like a Sunday-paper dating column rendered as software. Editorial type, single restrained accent, lots of negative space, no swipe deck or hookup tropes.
token for the metric numerals — these screens live or die on numerals.
real, specific copy. Examples: "the people who'd text back within a day", "manageable. two are now friends.", "your single greatest compatibility asset."
sans-serif eyebrow style: tagline left, "NEXT TIER AT 2,080 MUTUALS" right, both in mono caps with letter-spacing. Thin rule below.
top. User card (avatar / handle / ratio / tier). Three groups of nav: "TODAY" (specimen, inbox, queue, notifications), "YOU" (your stats, mutuals & communities, blocked, settings), "ARCHIVE" (past issues, expired matches). Active item gets accent text + accent dot.
caps mono label, an oversized serif numeral (use accent or muted green for positive, muted red for caution), one-line italic footnote. Plausible specifics — "1,842 ↑ 41 this wk · healthy growth.", "14% above median for your cohort.", "4 / exes in your circle · manageable. two are now friends."
bars, last two days highlighted in accent. Caption above with "↑ TRENDING UP · +3 CLOSE MUTUALS THIS MONTH · TWO VIA THE SAME OFFSITE" in mono.
copy below ("STEADY CLIMB FROM 8% → 14%. ATTRIBUTABLE TO ONE COMMUNITY JOIN…"). Footer rule.
<!doctype html> through </html>, CSS inline.font-feature-settings: 'tnum' on the metric numerals.data-od-id on ticker, sidebar, kpi grid, chart, trend.bars, one nav active state).
Emit between <artifact> tags:
<artifact identifier="dating-slug" type="text/html" title="Dating Dashboard — 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-05 | pass→fail | 26,499 | 30,364 | +15% | 1 | 1 | 0% | 6,174 | 6,961 | +13% | 0 | 0 | — |
case-19 | fail→fail | 28,088 | 28,046 | -0% | 1 | 1 | 0% | 6,178 | 6,965 | +13% | 0 | 0 | — |
case-20 | fail→pass | 25,591 | 27,511 | +8% | 1 | 1 | 0% | 6,179 | 6,966 | +13% | 0 | 0 | — |
case-01 | fail→fail | 31,998 | 29,086 | -9% | 1 | 1 | 0% | 6,208 | 6,995 | +13% | 0 | 0 | — |
case-02 | fail→fail | 30,250 | 44,415 | +47% | 1 | 1 | 0% | 6,191 | 6,978 | +13% | 0 | 0 | — |
case-03 | fail→fail | 29,711 | 53,497 | +80% | 1 | 1 | 0% | 6,188 | 6,975 | +13% | 0 | 0 | — |
case-04 | pass→fail | 26,001 | 28,187 | +8% | 1 | 1 | 0% | 6,183 | 6,970 | +13% | 0 | 0 | — |
case-06 | pass→fail | 15,018 | 27,857 | +85% | 1 | 1 | 0% | 3,034 | 6,950 | +129% | 0 | 0 | — |
case-07 | pass→pass | 35,683 | 27,579 | -23% | 1 | 1 | 0% | 6,193 | 6,980 | +13% | 0 | 0 | — |
case-08 | fail→fail | 28,614 | 27,257 | -5% | 1 | 1 | 0% | 6,201 | 6,988 | +13% | 0 | 0 | — |
case-09 | pass→pass | 28,630 | 25,370 | -11% | 1 | 1 | 0% | 6,189 | 6,976 | +13% | 0 | 0 | — |
case-10 | pass→fail | 26,457 | 25,207 | -5% | 1 | 1 | 0% | 6,186 | 6,973 | +13% | 0 | 0 | — |
case-11 | pass→fail | 19,737 | 25,442 | +29% | 1 | 1 | 0% | 4,345 | 6,970 | +60% | 0 | 0 | — |
case-12 | fail→fail | 28,514 | 31,245 | +10% | 1 | 1 | 0% | 6,190 | 6,977 | +13% | 0 | 0 | — |
case-13 | fail→fail | 30,265 | 26,277 | -13% | 1 | 1 | 0% | 6,184 | 6,971 | +13% | 0 | 0 | — |
case-14 | pass→fail | 13,698 | 27,070 | +98% | 1 | 1 | 0% | 2,828 | 6,983 | +147% | 0 | 0 | — |
case-15 | pass→pass | 29,691 | 27,845 | -6% | 1 | 1 | 0% | 6,193 | 6,979 | +13% | 0 | 0 | — |
case-16 | pass→pass | 23,813 | 33,006 | +39% | 1 | 1 | 0% | 6,181 | 6,968 | +13% | 0 | 0 | — |
case-17 | fail→pass | 25,352 | 26,289 | +4% | 1 | 1 | 0% | 6,184 | 6,971 | +13% | 0 | 0 | — |
case-18 | pass→pass | 22,528 | 27,275 | +21% | 1 | 1 | 0% | 4,832 | 6,969 | +44% | 0 | 0 | — |
case-21 | pass→pass | 24,860 | 27,540 | +11% | 1 | 1 | 0% | 5,927 | 6,967 | +18% | 0 | 0 | — |
case-22 | pass→pass | 23,085 | 26,545 | +15% | 1 | 1 | 0% | 4,754 | 6,971 | +47% | 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. 8 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.