Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Turn one piece of content into a full multi-platform pack — X/Twitter thread, LinkedIn post, newsletter section, Instagram carousel, and a short-form video script — each rewritten natively for its platform, not copy-pasted. Use when asked to repurpose content, atomize a blog post or video, turn one idea into many posts, or get more mileage from a piece. Produces ready-to-post drafts per platform with hooks, formatting, and CTAs tuned to each.
.claude/skills/itamarzand88-content-repurposer/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 41% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 59% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 141% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 240% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 113% | 0% |
<!-- source: content-repurposer — https://raw.githubusercontent.com/mohitagw15856/pm-claude-skills/main/plugins/pm-creator/skills/content-repurposer/SKILL.md -->
Creators don't have a content problem — they have a distribution problem. One good idea should become a week of posts. This skill atomizes a single source (a blog post, video transcript, newsletter, or raw notes) into platform-native drafts — each one rewritten for how people actually read on that platform, never just truncated.
Given a source (or a rough topic), produce the full pack anyway — pull the core insight and reshape it per platform. If the source is thin, extract the strongest single idea and build around it. Mark any invented stat/example (assumed — replace). Never output the same text five times with different line breaks.
Ask for (if not already provided):
Lead with The core idea in one sentence (everything else ladders to it). Then, per platform:
A scroll-stopping hook tweet, then 5–9 tweets each carrying one beat, a final CTA tweet. Tight, line-broken, no fluff.
A hook line + short-paragraph body (whitespace-heavy), a concrete takeaway, a soft CTA / question to drive comments. No hashtag spam (3–5 max).
A subject-line option, a one-line preview, and a 150–250-word section with a clear takeaway and link-out.
Slide 1 = the hook; slides 2–6 = one point each (≤12 words per slide + a sentence of body); final slide = CTA. Give the on-slide text and the caption.
A 0–3s hook line, the body beats with on-screen text cues, and a payoff/CTA. 30–45s of spoken copy.
End with:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-04 | fail→fail | 22,611 | 26,217 | +16% | 1 | 1 | 0% | 4,067 | 5,404 | +33% | 0 | 0 | — |
case-01 | pass→pass | 17,073 | 16,219 | -5% | 1 | 1 | 0% | 3,200 | 3,699 | +16% | 0 | 0 | — |
case-02 | pass→fail | 14,829 | 24,438 | +65% | 1 | 1 | 0% | 2,798 | 5,255 | +88% | 0 | 0 | — |
case-03 | fail→pass | 13,368 | 13,375 | +0% | 1 | 1 | 0% | 2,096 | 2,945 | +41% | 0 | 0 | — |
case-05 | fail→pass | 17,635 | 20,844 | +18% | 1 | 1 | 0% | 3,290 | 5,237 | +59% | 0 | 0 | — |
case-06 | fail→fail | 26,755 | 17,063 | -36% | 1 | 1 | 0% | 4,504 | 4,273 | -5% | 0 | 0 | — |
case-07 | pass→fail | 11,668 | 12,144 | +4% | 1 | 1 | 0% | 1,803 | 3,184 | +77% | 0 | 0 | — |
case-08 | fail→pass | 8,907 | 15,433 | +73% | 1 | 1 | 0% | 1,612 | 3,879 | +141% | 0 | 0 | — |
case-22 | pass→pass | 9,901 | 15,502 | +57% | 1 | 1 | 0% | 1,690 | 3,614 | +114% | 0 | 0 | — |
case-09 | pass→pass | 13,820 | 10,897 | -21% | 1 | 1 | 0% | 2,799 | 2,868 | +2% | 0 | 0 | — |
case-10 | fail→pass | 7,346 | 19,901 | +171% | 1 | 1 | 0% | 1,267 | 4,308 | +240% | 0 | 0 | — |
case-11 | fail→pass | 9,983 | 16,200 | +62% | 1 | 1 | 0% | 1,808 | 3,849 | +113% | 0 | 0 | — |
case-12 | fail→fail | 12,686 | 14,244 | +12% | 1 | 1 | 0% | 2,373 | 3,443 | +45% | 0 | 0 | — |
case-13 | fail→pass | 17,277 | 17,312 | +0% | 1 | 1 | 0% | 2,809 | 3,919 | +40% | 0 | 0 | — |
case-14 | fail→fail | 5,596 | 4,480 | -20% | 1 | 1 | 0% | 941 | 1,598 | +70% | 0 | 0 | — |
case-15 | pass→pass | 20,159 | 19,425 | -4% | 1 | 1 | 0% | 3,324 | 3,929 | +18% | 0 | 0 | — |
case-16 | fail→pass | 12,782 | 17,838 | +40% | 1 | 1 | 0% | 2,208 | 4,020 | +82% | 0 | 0 | — |
case-17 | pass→pass | 8,760 | 16,707 | +91% | 1 | 1 | 0% | 1,511 | 3,910 | +159% | 0 | 0 | — |
case-18 | pass→pass | 9,335 | 16,520 | +77% | 1 | 1 | 0% | 1,602 | 3,432 | +114% | 0 | 0 | — |
case-19 | fail→pass | 9,446 | 15,954 | +69% | 1 | 1 | 0% | 1,451 | 3,503 | +141% | 0 | 0 | — |
case-20 | pass→pass | 14,692 | 20,210 | +38% | 1 | 1 | 0% | 2,741 | 4,205 | +53% | 0 | 0 | — |
case-21 | pass→pass | 11,880 | 18,880 | +59% | 1 | 1 | 0% | 2,094 | 4,065 | +94% | 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 +27 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is 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.