Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →When the user wants to add, optimize, or audit popups or modals for lead capture or offers. Also use when the user mentions "popup," "modal," "lightbox," "overlay," "exit-intent," "popup form," "modal design," "lead popup," "popup timing," or "popup triggers." For CRO, use conversion-optimization.
.claude/skills/kostja94-popup-generator/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | -18% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -4% | 0% |
| case-20 | ✗→✓ | ▲ Improved | -39% | 0% |
| case-01 | ✓→✗ | ▼ Worse | 6% | 0% |
| case-12 | ✓→✓ | ＝ Same ✓ | -13% | 0% |
Guides popup and modal design for conversion. Well-designed popups can achieve up to 25% conversion; poorly timed or intrusive ones hurt UX and SEO. Google penalizes intrusive mobile popups.
When invoking: On first use, if helpful, open with 1-2 sentences on what this skill covers and why it matters, then provide the main output. On subsequent use or when the user asks to skip, go directly to the main output.
Check for project context first: If .claude/project-context.md or .cursor/project-context.md exists, read it for offers and audience.
Identify:
| Trigger | Use | |---------|-----| | Time delay | 5-15s typical; after some engagement | | Scroll % | 25-50% read; user invested | | Exit intent | Mouse leaving viewport; last chance | | Click | User-initiated; least intrusive |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-12 | pass→pass | 14,380 | 9,661 | -33% | 1 | 1 | 0% | 2,828 | 2,473 | -13% | 0 | 0 | — |
case-13 | pass→pass | 13,490 | 9,318 | -31% | 1 | 1 | 0% | 2,264 | 2,054 | -9% | 0 | 0 | — |
case-01 | pass→fail | 15,022 | 11,482 | -24% | 1 | 1 | 0% | 2,375 | 2,512 | +6% | 0 | 0 | — |
case-02 | fail→pass | 16,282 | 10,211 | -37% | 1 | 1 | 0% | 2,792 | 2,291 | -18% | 0 | 0 | — |
case-03 | fail→fail | 11,092 | 8,810 | -21% | 1 | 1 | 0% | 1,876 | 2,064 | +10% | 0 | 0 | — |
case-04 | pass→pass | 9,573 | 10,848 | +13% | 1 | 1 | 0% | 1,588 | 2,379 | +50% | 0 | 0 | — |
case-05 | fail→fail | 15,610 | 10,822 | -31% | 1 | 1 | 0% | 2,358 | 2,407 | +2% | 0 | 0 | — |
case-06 | pass→pass | 12,056 | 10,390 | -14% | 1 | 1 | 0% | 1,925 | 2,283 | +19% | 0 | 0 | — |
case-07 | fail→pass | 14,537 | 10,397 | -28% | 1 | 1 | 0% | 2,416 | 2,323 | -4% | 0 | 0 | — |
case-08 | pass→pass | 11,850 | 7,960 | -33% | 1 | 1 | 0% | 1,957 | 1,852 | -5% | 0 | 0 | — |
case-09 | pass→pass | 8,773 | 7,657 | -13% | 1 | 1 | 0% | 1,630 | 1,905 | +17% | 0 | 0 | — |
case-10 | pass→pass | 10,315 | 10,364 | +0% | 1 | 1 | 0% | 1,938 | 2,427 | +25% | 0 | 0 | — |
case-11 | pass→pass | 11,284 | 6,711 | -41% | 1 | 1 | 0% | 2,082 | 1,790 | -14% | 0 | 0 | — |
case-14 | pass→pass | 9,069 | 12,077 | +33% | 1 | 1 | 0% | 1,913 | 3,098 | +62% | 0 | 0 | — |
case-15 | pass→pass | 24,823 | 13,157 | -47% | 1 | 1 | 0% | 5,344 | 3,019 | -44% | 0 | 0 | — |
case-16 | pass→pass | 14,396 | 9,607 | -33% | 1 | 1 | 0% | 2,614 | 2,506 | -4% | 0 | 0 | — |
case-17 | pass→pass | 13,062 | 9,612 | -26% | 1 | 1 | 0% | 2,256 | 2,374 | +5% | 0 | 0 | — |
case-18 | pass→pass | 10,695 | 9,986 | -7% | 1 | 1 | 0% | 1,746 | 2,369 | +36% | 0 | 0 | — |
case-19 | pass→pass | 10,660 | 8,401 | -21% | 1 | 1 | 0% | 1,909 | 2,014 | +6% | 0 | 0 | — |
case-20 | fail→pass | 27,678 | 18,620 | -33% | 1 | 1 | 0% | 6,181 | 3,767 | -39% | 0 | 0 | — |
case-21 | pass→pass | 14,610 | 11,277 | -23% | 1 | 1 | 0% | 2,701 | 2,883 | +7% | 0 | 0 | — |
case-22 | pass→pass | 15,234 | 12,892 | -15% | 1 | 1 | 0% | 2,735 | 3,187 | +17% | 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 +9 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.