Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Project-specific UI refactor workflow for the Scoop news platform in /home/bender/classwork/Thesis. Use when updating Scoop frontend layouts, globe UX, research workspace, and brand docs while preserving all features and enforcing a no-emoji policy across UI, logs, and docs.
.claude/skills/aiskillstore-scoop-ui-refactor/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | 26% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -11% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -33% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -29% | 0% |
/home/bender/classwork/Thesisfrontend/app and frontend/components1) Review primary surfaces
frontend/app/page.tsx (home layout)frontend/components/globe-view.tsx and frontend/components/interactive-globe.tsxfrontend/app/search/page.tsx (research UI)frontend/app/globals.css (theme tokens)2) Home layout refactor
3) Globe view UX
4) Research workspace
5) No-emoji enforcement
6) Feed view snap-scroll (TikTok style)
h-screen, snap-start) within a scroll container that has snap-y snap-mandatory scroll-smooth.event.preventDefault() on wheel/touch when a snap transition actually occurs; allow default scrolling when at the first/last item.activeIndex and an isAnimating guard; call scrollIntoView for the next/previous section and clear the guard after a short timeout.h-screen without accounting for surrounding layout; prefer a flex parent (flex-1 h-full min-h-0) and per-section h-full min-h-full snap-start inside the scroll container so each article uses the available viewport height without hiding titles.any); fixes are pending and not related to snap-scroll layout.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 15,384 | 13,540 | -12% | 1 | 1 | 0% | 364 | 1,268 | +248% | 0 | 0 | — |
case-02 | fail→fail | 12,401 | 13,629 | +10% | 1 | 1 | 0% | 400 | 1,213 | +203% | 0 | 0 | — |
case-03 | fail→fail | 23,316 | 17,031 | -27% | 1 | 1 | 0% | 377 | 1,101 | +192% | 0 | 0 | — |
case-04 | pass→pass | 23,063 | 15,386 | -33% | 1 | 1 | 0% | 2,994 | 2,383 | -20% | 0 | 0 | — |
case-05 | fail→pass | 15,714 | 13,582 | -14% | 1 | 1 | 0% | 2,527 | 3,177 | +26% | 0 | 0 | — |
case-06 | fail→pass | 20,283 | 14,526 | -28% | 1 | 1 | 0% | 2,113 | 1,874 | -11% | 0 | 0 | — |
case-07 | fail→fail | 17,086 | 4,845 | -72% | 1 | 1 | 0% | 1,894 | 1,361 | -28% | 0 | 0 | — |
case-08 | fail→pass | 12,715 | 11,017 | -13% | 1 | 1 | 0% | 1,940 | 1,688 | -13% | 0 | 0 | — |
case-09 | fail→pass | 22,296 | 10,031 | -55% | 1 | 1 | 0% | 3,023 | 2,037 | -33% | 0 | 0 | — |
case-10 | pass→pass | 18,689 | 9,637 | -48% | 1 | 1 | 0% | 2,111 | 1,375 | -35% | 0 | 0 | — |
case-11 | fail→pass | 31,598 | 11,205 | -65% | 1 | 1 | 0% | 2,325 | 1,662 | -29% | 0 | 0 | — |
case-12 | fail→fail | 18,506 | 16,676 | -10% | 1 | 1 | 0% | 2,016 | 1,035 | -49% | 0 | 0 | — |
case-13 | pass→pass | 15,684 | 9,873 | -37% | 1 | 1 | 0% | 2,293 | 1,491 | -35% | 0 | 0 | — |
case-14 | pass→pass | 19,026 | 14,592 | -23% | 1 | 1 | 0% | 2,056 | 2,071 | +1% | 0 | 0 | — |
case-15 | pass→pass | 24,471 | 8,241 | -66% | 1 | 1 | 0% | 2,721 | 1,191 | -56% | 0 | 0 | — |
case-16 | fail→pass | 37,603 | 12,933 | -66% | 1 | 1 | 0% | 2,207 | 1,899 | -14% | 0 | 0 | — |
case-17 | fail→pass | 24,246 | 19,708 | -19% | 1 | 1 | 0% | 1,591 | 1,323 | -17% | 0 | 0 | — |
case-18 | pass→pass | 32,974 | 12,375 | -62% | 1 | 1 | 0% | 2,346 | 1,595 | -32% | 0 | 0 | — |
case-19 | fail→fail | 54,541 | 15,395 | -72% | 1 | 1 | 0% | 2,122 | 1,163 | -45% | 0 | 0 | — |
case-20 | fail→fail | 36,552 | 21,071 | -42% | 1 | 1 | 0% | 2,200 | 985 | -55% | 0 | 0 | — |
case-21 | pass→fail | 37,553 | 36,958 | -2% | 1 | 1 | 0% | 4,009 | 919 | -77% | 0 | 0 | — |
case-22 | pass→fail | 14,275 | 76,284 | +434% | 1 | 1 | 0% | 1,655 | 913 | -45% | 0 | 0 | — |
case-23 | pass→pass | 41,395 | 59,544 | +44% | 1 | 1 | 0% | 3,198 | 4,664 | +46% | 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 17 counted toward the lift figure. The other 6 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 +22 percentage points is the difference between those two pass rates over the 17 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.