Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →When the user wants to create or optimize a refund or return policy page. Also use when the user mentions "refund policy," "return policy," "money-back guarantee," "returns and refunds," "refund page," "return process," "refund terms," or "satisfaction guarantee." For legal overview, use legal-page-generator.
.claude/skills/sediman-agent-refund-page-generator/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 4% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -1% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 7% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 37% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -3% | 0% |
Guides refund and return policy page content for e-commerce and digital products.
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.
Identify:
| Element | Purpose | |---------|---------| | Eligibility | What can be returned; time limits | | Process | How to request; steps | | Refund method | Original payment; timeline | | Exclusions | Non-refundable items | | Contact | Support for returns |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 18,602 | 14,264 | -23% | 1 | 1 | 0% | 2,941 | 2,642 | -10% | 0 | 0 | — |
case-02 | fail→pass | 12,872 | 11,206 | -13% | 1 | 1 | 0% | 1,904 | 1,979 | +4% | 0 | 0 | — |
case-03 | fail→pass | 17,716 | 15,273 | -14% | 1 | 1 | 0% | 2,786 | 2,749 | -1% | 0 | 0 | — |
case-04 | fail→fail | 15,882 | 10,268 | -35% | 1 | 1 | 0% | 2,196 | 1,843 | -16% | 0 | 0 | — |
case-05 | fail→pass | 14,268 | 13,351 | -6% | 1 | 1 | 0% | 2,121 | 2,265 | +7% | 0 | 0 | — |
case-06 | fail→fail | 13,971 | 10,932 | -22% | 1 | 1 | 0% | 2,152 | 2,033 | -6% | 0 | 0 | — |
case-07 | fail→pass | 7,934 | 8,255 | +4% | 1 | 1 | 0% | 1,177 | 1,613 | +37% | 0 | 0 | — |
case-08 | fail→pass | 14,822 | 12,337 | -17% | 1 | 1 | 0% | 2,157 | 2,103 | -3% | 0 | 0 | — |
case-09 | fail→pass | 13,459 | 13,129 | -2% | 1 | 1 | 0% | 1,992 | 2,329 | +17% | 0 | 0 | — |
case-10 | fail→pass | 15,465 | 12,481 | -19% | 1 | 1 | 0% | 2,189 | 2,056 | -6% | 0 | 0 | — |
case-11 | fail→fail | 16,019 | 12,951 | -19% | 1 | 1 | 0% | 2,388 | 2,249 | -6% | 0 | 0 | — |
case-12 | fail→pass | 16,774 | 15,321 | -9% | 1 | 1 | 0% | 2,620 | 2,617 | -0% | 0 | 0 | — |
case-13 | fail→pass | 22,009 | 15,744 | -28% | 1 | 1 | 0% | 3,099 | 2,631 | -15% | 0 | 0 | — |
case-14 | fail→pass | 16,686 | 13,929 | -17% | 1 | 1 | 0% | 2,401 | 2,417 | +1% | 0 | 0 | — |
case-15 | fail→pass | 13,249 | 12,139 | -8% | 1 | 1 | 0% | 1,858 | 2,050 | +10% | 0 | 0 | — |
case-16 | fail→pass | 16,978 | 12,012 | -29% | 1 | 1 | 0% | 2,308 | 2,103 | -9% | 0 | 0 | — |
case-17 | pass→pass | 14,571 | 11,016 | -24% | 1 | 1 | 0% | 2,290 | 2,056 | -10% | 0 | 0 | — |
case-18 | fail→pass | 20,982 | 12,504 | -40% | 1 | 1 | 0% | 2,950 | 1,995 | -32% | 0 | 0 | — |
case-19 | pass→pass | 16,973 | 14,556 | -14% | 1 | 1 | 0% | 3,009 | 2,751 | -9% | 0 | 0 | — |
case-20 | pass→pass | 23,154 | 13,943 | -40% | 1 | 1 | 0% | 3,839 | 2,698 | -30% | 0 | 0 | — |
case-21 | pass→pass | 10,705 | 9,878 | -8% | 1 | 1 | 0% | 1,695 | 1,813 | +7% | 0 | 0 | — |
case-22 | fail→pass | 17,710 | 14,640 | -17% | 1 | 1 | 0% | 2,555 | 2,413 | -6% | 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 +64 percentage points is the difference between those two pass rates over the 22 comparable cases.
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.