Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when the user wants to design, package, or improve an offer — the thing being sold, not just the price or copy. Trigger on 'offer', 'irresistible offer', 'value stack', 'bonus stack', 'guarantee', 'risk reversal', 'scarcity', 'urgency', 'upsell', 'downsell', 'high-ticket', 'productized service', 'coaching offer', 'agency offer', or 'offer is not converting'.
.claude/skills/minhnv0807-31-offer-design-global/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | -14% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -36% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 68% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 9% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 68% | 0% |
An offer is what the customer actually buys: outcome, package, proof, guarantee, bonuses, constraints, and payment structure. Better copy cannot rescue a weak offer.
Read .agents/product-marketing-context-global.md if available. Ask up to 4 questions: what is sold, who buys it, current or intended price, and why buyers hesitate. If the user has a sales page, proposal, pricing sheet, or funnel, use it as input. For details, read references/offer-framework.md.
Use this skill for:
If the user only needs price tiers, margins, or break-even math, use 17-pricing-strategy-global first.
Score each dimension 1-5:
| Dimension | Question | |-----------|----------| | Outcome | What concrete result does the buyer get? | | Proof | Why should they believe this will work? | | Speed | How soon do they see first value? | | Effort | How much must the buyer do alone? | | Risk | Who carries risk if it fails? |
Fix the lowest score first.
Use 4 levers:
Include:
When useful, create 3 levels:
| Level | Role | |-------|------| | Entry | Lower friction and build trust | | Core | Main offer to sell most often | | Premium | Faster, done-for-you, or higher-touch support |
Remove fake scarcity, impossible guarantees, inflated bonus value, exaggerated claims, and get-rich-fast language.
markdown# Offer Design — [Offer name] ## 1. Current diagnosis | Component | Score | Problem | Priority fix | ## 2. Buyer and outcome - ICP: - Pain: - Dream outcome: - First value moment: ## 3. New offer - Name: - Core deliverable: - Bonuses: - Guarantee: - Real scarcity/urgency: - Price/payment: ## 4. Value stack | Component | Objection handled | Perceived value | Notes | ## 5. Funnel implications - Landing page changes: - Ad/copy angles: - Email follow-up: - Upsell/downsell: ## 6. Test plan | Hypothesis | Variable | Metric | Duration |
17-pricing-strategy-global: price tiers, break-even, and pricing psychology.05-ad-copy-global: copy that presents the offer.12-landing-page-brief-global: landing page structure.19-ab-test-setup-global: testing guarantee, bonus, headline, or price display.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 35,693 | 32,333 | -9% | 1 | 1 | 0% | 5,743 | 5,519 | -4% | 0 | 0 | — |
case-02 | fail→pass | 36,590 | 21,841 | -40% | 1 | 1 | 0% | 5,259 | 4,519 | -14% | 0 | 0 | — |
case-03 | fail→pass | 41,897 | 26,381 | -37% | 1 | 1 | 0% | 7,412 | 4,713 | -36% | 0 | 0 | — |
case-04 | pass→fail | 12,807 | 14,856 | +16% | 1 | 1 | 0% | 2,830 | 4,009 | +42% | 0 | 0 | — |
case-05 | pass→pass | 10,805 | 19,240 | +78% | 1 | 1 | 0% | 2,404 | 4,006 | +67% | 0 | 0 | — |
case-06 | pass→pass | 20,798 | 21,773 | +5% | 1 | 1 | 0% | 2,443 | 3,572 | +46% | 0 | 0 | — |
case-07 | pass→pass | 25,151 | 27,348 | +9% | 1 | 1 | 0% | 3,369 | 4,641 | +38% | 0 | 0 | — |
case-08 | pass→pass | 17,957 | 20,170 | +12% | 1 | 1 | 0% | 2,976 | 4,133 | +39% | 0 | 0 | — |
case-09 | pass→pass | 16,896 | 18,860 | +12% | 1 | 1 | 0% | 2,979 | 4,067 | +37% | 0 | 0 | — |
case-10 | pass→pass | 19,356 | 20,583 | +6% | 1 | 1 | 0% | 3,204 | 4,408 | +38% | 0 | 0 | — |
case-11 | fail→pass | 15,191 | 22,093 | +45% | 1 | 1 | 0% | 2,724 | 4,585 | +68% | 0 | 0 | — |
case-12 | fail→fail | 12,535 | 7,661 | -39% | 1 | 1 | 0% | 2,284 | 2,300 | +1% | 0 | 0 | — |
case-13 | fail→pass | 9,312 | 4,450 | -52% | 1 | 1 | 0% | 1,513 | 1,649 | +9% | 0 | 0 | — |
case-14 | pass→pass | 22,315 | 19,720 | -12% | 1 | 1 | 0% | 3,787 | 4,278 | +13% | 0 | 0 | — |
case-15 | fail→pass | 12,844 | 15,566 | +21% | 1 | 1 | 0% | 2,064 | 3,467 | +68% | 0 | 0 | — |
case-16 | pass→pass | 10,435 | 7,549 | -28% | 1 | 1 | 0% | 1,952 | 2,271 | +16% | 0 | 0 | — |
case-17 | pass→pass | 7,631 | 7,903 | +4% | 1 | 1 | 0% | 1,195 | 2,244 | +88% | 0 | 0 | — |
case-18 | pass→pass | 16,303 | 11,038 | -32% | 1 | 1 | 0% | 2,794 | 2,668 | -5% | 0 | 0 | — |
case-19 | pass→pass | 17,752 | 22,147 | +25% | 1 | 1 | 0% | 3,004 | 4,453 | +48% | 0 | 0 | — |
case-20 | fail→pass | 25,602 | 29,024 | +13% | 1 | 1 | 0% | 4,100 | 5,616 | +37% | 0 | 0 | — |
case-21 | pass→pass | 18,069 | 18,289 | +1% | 1 | 1 | 0% | 2,976 | 4,343 | +46% | 0 | 0 | — |
case-22 | pass→pass | 14,374 | 21,005 | +46% | 1 | 1 | 0% | 2,334 | 4,420 | +89% | 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 +23 percentage points is the difference between those two pass rates over the 22 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.