Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Write a product description / listing that sells and ranks. Use when asked to write a product description, e-commerce listing copy, a product page, or to rewrite a flat product blurb. Produces benefit-led listing copy — a hook, scannable feature→benefit bullets, specs, an SEO-aware title and keywords, and trust/again-objection elements — tuned to the buyer and channel.
.claude/skills/mohitagw15856-product-description/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 45% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 4% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 45% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 32% | 0% |
Shoppers skim, then decide. A product description wins when it leads with the benefit (what changes for the buyer), makes the value scannable, and answers the objection that would stop the "add to cart" — while weaving in the search terms people actually type. This skill turns a spec sheet into copy that sells and gets found.
Given just a product name and a few features, write the full listing anyway — infer the buyer, the benefits, and likely keywords from the product type, and mark anything inferred (assumed — confirm). Never invent specs, materials, certifications, or claims (especially health/safety/efficacy) — leave those bracketed to confirm. Never hand back questions instead of copy.
Ask for these only if they aren't already provided (else infer and label):
Mark any inferred spec/claim (assumed — confirm).
E-commerce copywriting practice — benefit-led, scannable listings with feature-to-benefit translation, on-page SEO, and objection handling.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 21,045 | 20,912 | -1% | 1 | 1 | 0% | 2,332 | 3,391 | +45% | 0 | 0 | — |
case-02 | fail→pass | 23,637 | 18,462 | -22% | 1 | 1 | 0% | 3,251 | 3,372 | +4% | 0 | 0 | — |
case-03 | fail→pass | 21,082 | 16,785 | -20% | 1 | 1 | 0% | 2,669 | 2,961 | +11% | 0 | 0 | — |
case-04 | pass→pass | 18,453 | 23,108 | +25% | 1 | 1 | 0% | 2,337 | 3,971 | +70% | 0 | 0 | — |
case-05 | pass→pass | 15,870 | 17,150 | +8% | 1 | 1 | 0% | 2,205 | 3,353 | +52% | 0 | 0 | — |
case-06 | pass→pass | 24,967 | 26,185 | +5% | 1 | 1 | 0% | 2,912 | 4,767 | +64% | 0 | 0 | — |
case-07 | fail→pass | 17,125 | 16,410 | -4% | 1 | 1 | 0% | 1,954 | 2,834 | +45% | 0 | 0 | — |
case-08 | fail→fail | 6,881 | 18,127 | +163% | 1 | 1 | 0% | 1,201 | 2,581 | +115% | 0 | 0 | — |
case-09 | pass→pass | 13,935 | 15,210 | +9% | 1 | 1 | 0% | 1,278 | 2,551 | +100% | 0 | 0 | — |
case-10 | fail→fail | 23,059 | 19,211 | -17% | 1 | 1 | 0% | 3,078 | 3,136 | +2% | 0 | 0 | — |
case-11 | fail→pass | 21,598 | 17,433 | -19% | 1 | 1 | 0% | 2,220 | 2,928 | +32% | 0 | 0 | — |
case-12 | fail→fail | 17,354 | 18,690 | +8% | 1 | 1 | 0% | 2,017 | 3,032 | +50% | 0 | 0 | — |
case-13 | fail→fail | 19,654 | 17,281 | -12% | 1 | 1 | 0% | 2,284 | 2,729 | +19% | 0 | 0 | — |
case-14 | fail→fail | 18,335 | 14,888 | -19% | 1 | 1 | 0% | 2,171 | 2,878 | +33% | 0 | 0 | — |
case-15 | pass→pass | 20,990 | 23,197 | +11% | 1 | 1 | 0% | 2,043 | 3,176 | +55% | 0 | 0 | — |
case-16 | fail→pass | 15,814 | 14,248 | -10% | 1 | 1 | 0% | 2,542 | 3,185 | +25% | 0 | 0 | — |
case-17 | fail→pass | 18,773 | 19,135 | +2% | 1 | 1 | 0% | 2,521 | 3,244 | +29% | 0 | 0 | — |
case-18 | fail→pass | 22,465 | 21,482 | -4% | 1 | 1 | 0% | 2,575 | 2,977 | +16% | 0 | 0 | — |
case-19 | fail→fail | 14,241 | 12,922 | -9% | 1 | 1 | 0% | 2,037 | 2,563 | +26% | 0 | 0 | — |
case-20 | fail→pass | 22,951 | 19,796 | -14% | 1 | 1 | 0% | 2,526 | 2,831 | +12% | 0 | 0 | — |
case-21 | fail→fail | 21,474 | 18,682 | -13% | 1 | 1 | 0% | 2,269 | 2,910 | +28% | 0 | 0 | — |
case-22 | fail→pass | 22,284 | 11,345 | -49% | 1 | 1 | 0% | 2,530 | 2,634 | +4% | 0 | 0 | — |
case-23 | fail→fail | 21,205 | 18,099 | -15% | 1 | 1 | 0% | 2,566 | 3,170 | +24% | 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. The headline lift of +43 percentage points is the difference between those two pass rates over the 23 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.