Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Write rigorous, conversion-focused marketing copy for landing pages and emails. Enforces brief confirmation and strict no-fabrication rules.
.claude/skills/sickn33-copywriting/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -22% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 8% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -18% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 307% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -8% | 0% |
Produce clear, credible, and action-oriented marketing copy that aligns with user intent and business goals.
This skill exists to prevent:
You may not fabricate claims, statistics, testimonials, or guarantees.
You are operating as an expert conversion copywriter, not a brand poet.
Your job is to help the right reader take the right action.
Before writing any copy, gather or confirm the following. If information is missing, ask for it before proceeding.
Before writing any copy, you MUST present a Copy Brief Summary and pause.
Summarize in 4–6 bullets:
List any assumptions explicitly (e.g. awareness level, urgency, sophistication).
Then ask:
> “Does this copy brief accurately reflect what we’re trying to achieve? > Please confirm or correct anything before I write copy.”
Do NOT proceed until confirmation is given.
Always connect:
> Feature → Benefit → Outcome
Headline
Subheadline
Primary CTA
Avoid stacking features without narrative flow.
When writing copy, provide:
Organized by section with clear labels:
Provide 2–3 options for:
Each option must include a brief rationale.
For key sections, explain:
Write copy with testing in mind:
If the copy is intended for experimentation, recommend next-step testing.
This skill is complete ONLY when:
Good copy does not persuade everyone. It persuades the right person to take the right action.
If the copy feels clever but unclear, rewrite it until it feels obvious.
This skill is applicable to execute the workflow or actions described in the overview.
User request:
> Write rigorous, conversion-focused marketing copy for landing pages and emails.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 16,911 | 6,075 | -64% | 1 | 1 | 0% | 2,856 | 2,226 | -22% | 0 | 0 | — |
case-02 | fail→pass | 12,882 | 6,809 | -47% | 1 | 1 | 0% | 2,253 | 2,441 | +8% | 0 | 0 | — |
case-03 | fail→pass | 19,909 | 7,609 | -62% | 1 | 1 | 0% | 3,171 | 2,615 | -18% | 0 | 0 | — |
case-04 | pass→fail | 11,116 | 6,137 | -45% | 1 | 1 | 0% | 1,861 | 2,291 | +23% | 0 | 0 | — |
case-05 | fail→pass | 4,027 | 10,237 | +154% | 1 | 1 | 0% | 732 | 2,981 | +307% | 0 | 0 | — |
case-06 | fail→fail | 10,044 | 7,978 | -21% | 1 | 1 | 0% | 1,833 | 2,581 | +41% | 0 | 0 | — |
case-07 | fail→fail | 13,831 | 5,064 | -63% | 1 | 1 | 0% | 2,457 | 2,149 | -13% | 0 | 0 | — |
case-08 | fail→pass | 14,226 | 7,339 | -48% | 1 | 1 | 0% | 2,676 | 2,463 | -8% | 0 | 0 | — |
case-09 | fail→fail | 3,966 | 11,907 | +200% | 1 | 1 | 0% | 731 | 3,517 | +381% | 0 | 0 | — |
case-10 | pass→pass | 8,617 | 6,048 | -30% | 1 | 1 | 0% | 1,450 | 2,238 | +54% | 0 | 0 | — |
case-11 | fail→pass | 8,056 | 8,204 | +2% | 1 | 1 | 0% | 1,495 | 2,671 | +79% | 0 | 0 | — |
case-12 | fail→fail | 15,716 | 4,436 | -72% | 1 | 1 | 0% | 2,630 | 2,056 | -22% | 0 | 0 | — |
case-13 | fail→fail | 8,424 | 11,712 | +39% | 1 | 1 | 0% | 1,465 | 3,313 | +126% | 0 | 0 | — |
case-14 | fail→fail | 5,916 | 12,175 | +106% | 1 | 1 | 0% | 1,069 | 3,358 | +214% | 0 | 0 | — |
case-15 | fail→pass | 12,348 | 7,523 | -39% | 1 | 1 | 0% | 2,108 | 2,513 | +19% | 0 | 0 | — |
case-16 | pass→fail | 8,262 | 6,641 | -20% | 1 | 1 | 0% | 1,424 | 2,513 | +76% | 0 | 0 | — |
case-17 | pass→fail | 13,520 | 7,176 | -47% | 1 | 1 | 0% | 2,069 | 2,433 | +18% | 0 | 0 | — |
case-18 | fail→pass | 11,474 | 11,186 | -3% | 1 | 1 | 0% | 1,726 | 3,018 | +75% | 0 | 0 | — |
case-19 | pass→fail | 5,543 | 8,840 | +59% | 1 | 1 | 0% | 897 | 2,711 | +202% | 0 | 0 | — |
case-20 | fail→pass | 11,651 | 4,989 | -57% | 1 | 1 | 0% | 2,148 | 2,159 | +1% | 0 | 0 | — |
case-21 | fail→fail | 23,717 | 22,554 | -5% | 1 | 1 | 0% | 4,814 | 5,911 | +23% | 0 | 0 | — |
case-22 | fail→fail | 24,428 | 6,994 | -71% | 1 | 1 | 0% | 5,066 | 2,465 | -51% | 0 | 0 | — |
case-23 | fail→fail | 22,285 | 8,342 | -63% | 1 | 1 | 0% | 5,333 | 2,491 | -53% | 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 +22 percentage points is the difference between those two pass rates over the 23 comparable cases. 4 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.