Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Integrate Stripe, PayPal, and payment processors. Handles checkout flows, subscriptions, webhooks, and PCI compliance. Use PROACTIVELY when implementing payments, billing, or subscription features.
.claude/skills/sickn33-payment-integration/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-20 | ✗→✓ | ▲ Improved | 64% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 47% | 0% |
| case-10 | ✓→✗ | ▼ Worse | 35% | 0% |
| case-17 | ✓→✗ | ▼ Worse | 62% | 0% |
| case-21 | ✓→✗ | ▼ Worse | 67% | 0% |
resources/implementation-playbook.md.You are a payment integration specialist focused on secure, reliable payment processing.
2xx status within 200ms, BEFORE expensive operations (database writes, external APIs). Timeouts trigger retries and duplicate processing.Real-world examples from Stripe, PayPal, OWASP:
Sources: Stripe official docs, PayPal Security Guidelines, OWASP Testing Guide, production retrospectives
Always use official SDKs. Include both server-side and client-side code where needed.
User request:
> Integrate Stripe, PayPal, and payment processors.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-20 | fail→pass | 11,779 | 14,919 | +27% | 1 | 1 | 0% | 2,477 | 4,060 | +64% | 0 | 0 | — |
case-05 | pass→pass | 14,834 | 11,176 | -25% | 1 | 1 | 0% | 3,266 | 2,974 | -9% | 0 | 0 | — |
case-01 | fail→fail | 26,224 | 20,812 | -21% | 1 | 1 | 0% | 6,018 | 5,442 | -10% | 0 | 0 | — |
case-02 | fail→fail | 26,654 | 26,021 | -2% | 1 | 1 | 0% | 6,200 | 6,812 | +10% | 0 | 0 | — |
case-03 | fail→fail | 22,438 | 21,823 | -3% | 1 | 1 | 0% | 5,155 | 5,490 | +6% | 0 | 0 | — |
case-04 | pass→pass | 20,781 | 23,896 | +15% | 1 | 1 | 0% | 4,333 | 5,864 | +35% | 0 | 0 | — |
case-06 | pass→pass | 15,450 | 17,373 | +12% | 1 | 1 | 0% | 3,096 | 3,951 | +28% | 0 | 0 | — |
case-07 | pass→pass | 11,949 | 13,693 | +15% | 1 | 1 | 0% | 2,371 | 3,756 | +58% | 0 | 0 | — |
case-08 | pass→pass | 21,104 | 16,839 | -20% | 1 | 1 | 0% | 3,930 | 4,188 | +7% | 0 | 0 | — |
case-09 | fail→pass | 14,023 | 16,034 | +14% | 1 | 1 | 0% | 2,606 | 3,824 | +47% | 0 | 0 | — |
case-10 | pass→fail | 14,161 | 12,451 | -12% | 1 | 1 | 0% | 2,397 | 3,232 | +35% | 0 | 0 | — |
case-11 | pass→pass | 12,048 | 13,045 | +8% | 1 | 1 | 0% | 2,217 | 3,352 | +51% | 0 | 0 | — |
case-12 | pass→pass | 14,440 | 14,971 | +4% | 1 | 1 | 0% | 2,669 | 3,670 | +38% | 0 | 0 | — |
case-13 | pass→pass | 16,292 | 14,129 | -13% | 1 | 1 | 0% | 3,090 | 3,282 | +6% | 0 | 0 | — |
case-14 | pass→pass | 10,674 | 9,127 | -14% | 1 | 1 | 0% | 1,949 | 2,333 | +20% | 0 | 0 | — |
case-15 | pass→pass | 11,911 | 15,796 | +33% | 1 | 1 | 0% | 2,097 | 3,727 | +78% | 0 | 0 | — |
case-16 | pass→pass | 16,075 | 16,067 | -0% | 1 | 1 | 0% | 3,257 | 3,927 | +21% | 0 | 0 | — |
case-17 | pass→fail | 12,613 | 16,001 | +27% | 1 | 1 | 0% | 2,441 | 3,960 | +62% | 0 | 0 | — |
case-18 | pass→pass | 14,931 | 17,848 | +20% | 1 | 1 | 0% | 2,904 | 4,221 | +45% | 0 | 0 | — |
case-19 | pass→pass | 11,915 | 11,139 | -7% | 1 | 1 | 0% | 2,311 | 2,866 | +24% | 0 | 0 | — |
case-21 | pass→fail | 11,188 | 15,100 | +35% | 1 | 1 | 0% | 2,254 | 3,768 | +67% | 0 | 0 | — |
case-22 | fail→fail | 14,765 | 14,477 | -2% | 1 | 1 | 0% | 2,559 | 3,290 | +29% | 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 -5 percentage points is the difference between those two pass rates over the 22 comparable cases. 3 cases got worse with the skill loaded, and they are included in that figure.
The publisher has shipped newer versions since this run, so these numbers describe v1, not the version currently listed.
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.