content-atomizer
Skill30 starsChanged 5 months agoMITView on GitHub
Turn one piece of long-form content (blog post, video/podcast transcript, guide, talk) into a ready-to-post multi-platform content pack: LinkedIn posts, X/Twitter thread, newsletter item, short-video scripts, carousel outlines, quote pulls, and more — each natively adapted, not copy-pasted. Use when the user says 'repurpose this post', 'turn this into social content', 'atomize this', 'get more mileage out of my content', 'make LinkedIn posts from my article', or shares long-form content wanting distribution. Produces an idea inventory plus a per-platform atomization pack with a suggested posting order.
3 other projects use a version of this.
Adopted by
Most recently taken up by
- MoizIbnYousaf/marketing-clitheir own version, first seen Nov 1, 2026
- genfeedai/genfeed.aitheir own version, first seen Sep 12, 2026
- davekindl/skillstheir own version, first seen Jul 23, 2026
Other versions3 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 3
- MoizIbnYousaf/marketing-cli★ 33same name
- genfeedai/genfeed.ai★ 4same name
- davekindl/skills★ 2same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository39 other files it gives its agents.
- brand-assetSkill
- creative-strategistSkill
- image-generationSkill
- orchestratorSkill
- original-designSkill
- product-photographySkill
- product-videoSkill
- remotion-script-writerSkill
- social-graphicsSkill
- talking-headSkill
- accessibility-excellenceSkill
- color-systemSkill
- component-architectureSkill
- design-engineer-mindsetSkill
- design-foundationSkill
- error-handling-recoverySkill
- frontend-orchestratorSkill
- interaction-physicsSkill
- layout-systemSkill
- loading-statesSkill
- performance-optimizationSkill
- typography-systemSkill
- visual-hierarchy-refactoringSkill
- brand-voiceSkill
- direct-response-copySkill
- email-sequencesSkill
- keyword-researchSkill
- lead-magnetSkill
- newsletterSkill
- orchestratorSkill
- positioning-anglesSkill
- seo-contentSkill
- seo-strategySkill
- tweet-writerSkill
- product-design-reviewSkill
- fable-like-implementationSkill
- foundation-sprintSkill
- sol-orchestratorSkill
- sprint-runnerSkill
The file
We split this skill into the rules it gives an agent
10 rules. About 0 are common, 7 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈7
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 12
Use when there is existing long-form source content and the goal is distribution across channels.
- Very specific: line 13
If the user needs the long-form piece written first, use skills/marketing/seo-content — then come back.
- Very specific: line 14
If the user wants one great tweet/thread crafted and researched from scratch (not derived from source content), use …
- Very specific: line 15
If the user wants a full newsletter issue (not just an item), hand the source to skills/marketing/newsletter.
- Distinctive: line 19
Ask in one batch, only what's missing:
- Distinctive: line 26
Don't ask which formats they want — that's your job to propose in Step 3.
- Distinctive: line 32
Read the seed content and pull 5–10 discrete ideas, each tagged by type — the type determines which formats it can power:
- Distinctive: line 41
If you can't find at least 4 distinct ideas, the content is too thin for a full pack — tell the user and scale scope down.
- Distinctive: line 73
Keep facts and claims within what the seed content supports — atomizing must not amplify a hedge into a certainty or invent numbers.
- Distinctive: line 119
Check every piece before delivering:
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Content Atomizer
Break one substantial piece of content into many platform-native pieces. Prime directive: adapt, never excerpt — each atomic piece must stand alone as if written for that platform on purpose, carrying the user's voice, while format and length obey the platform. The second rule: atomize ideas, not paragraphs — you extract the content's core ideas first, then re-express each idea per platform.
When to use / when not to
- Use when there is existing long-form source content and the goal is distribution across channels.
- If the user needs the long-form piece written first, use
skills/marketing/seo-content— then come back. - If the user wants one great tweet/thread crafted and researched from scratch (not derived from source content), use
skills/marketing/tweet-writer— it does niche research this skill skips. - If the user wants a full newsletter issue (not just an item), hand the source to
skills/marketing/newsletter.
Intake
Ask in one batch, only what's missing:
- The seed content — the actual text/transcript (or link). Verify it's substantial enough: multiple distinct ideas and a point of view. If it's thin (one idea, <~500 words), say so and offer a single-platform adaptation instead of a pack.
- Active platforms — where does the user actually publish, and which 1–2 matter most?
- Goal — reach/followers, traffic back to the source, or list growth? (This sets the CTAs.)
- Voice — brand-voice guide or sample posts (use
skills/marketing/brand-voiceoutput if it exists).
Don't ask which formats they want — that's your job to propose in Step 3. If platforms are unknown, default to LinkedIn + X + newsletter item and say so. State assumptions and proceed — don't stall.
Workflow
1. Extract the idea inventory
Read the seed content and pull 5–10 discrete ideas, each tagged by type — the type determines which formats it can power:
- Core concept — the piece's main argument
- Framework/process — anything with steps or named parts (best carousel/thread material)
- Story/example — narratives and cases (best LinkedIn/newsletter material)
- Surprising insight or contrarian take — the counterintuitive bits (best hook material)
- Actionable tip — standalone do-this-today advice (best short-video/single-post material)
- Quotable line — sentences that sting (quote graphics, thread closers)
If you can't find at least 4 distinct ideas, the content is too thin for a full pack — tell the user and scale scope down.
2. Select formats
Pick 5–7 formats, never all 15 — quality over coverage (heuristic: a strong pack on the user's 2 priority platforms beats a thin spray across 8). Decision rules:
- B2B / professional audience → weight LinkedIn posts + carousel; consumer/creator audience → weight short-video scripts + X.
- Goal = traffic/list growth → include the newsletter item and give posts link-out CTAs; goal = reach → engagement CTAs, links in replies/comments.
- Framework-type ideas present → include a carousel or thread; strong story present → include a story-post; no visual asset capacity → skip infographic/quote graphics, deliver text specs only.
Format spec table (length and structure per format):
| Format | Yield from one seed | Spec |
| LinkedIn posts | 3–5 | 100–200 words; hook line → story/insight → lesson → question or CTA; line breaks for scannability |
| X/Twitter thread | 1–2 | 5–10 tweets; hook tweet ≤250 chars; one idea per tweet; closer with CTA (full craft: skills/marketing/tweet-writer) |
| Single tweets | 3–5 | ≤110 chars preferred; sharpest standalone insights |
| Newsletter item | 1 | 300–500 words; hook → insight → lesson → link to full piece |
| Short-video scripts (TikTok/Reels/Shorts) | 2–3 | 30–60s; hook (≤3s) → problem → insight → CTA; write as spoken lines with [B-ROLL] notes |
| Carousel outline | 1–2 | 5–10 slides; slide 1 = hook, one point per slide, last slide = CTA; text per slide ≤25 words |
| Quote graphics (specs) | 3–5 | One pulled quote + attribution; supply exact text, no design |
| Instagram captions | 2–3 | 100–150 words; hook first line (feed truncates), story → lesson → CTA |
| Slide deck outline | 1 | 10–15 slides: title → problem → 3–5 content → conclusion → CTA |
| FAQ/explainer | 1 | 5–10 Q&As, 40–80 word answers (feeds SEO FAQ blocks) |
| Case study | 1 | 1,000–1,500 words; situation → action → result → lesson (only if seed contains a real result) |
| Webinar/talk outline | 1 | Timed outline: intro → problem → solution → examples → Q&A → CTA (only on request) |
3. Propose the pack, then generate
Show the user a one-screen plan first: idea inventory + chosen formats + which idea powers which piece. Get a quick confirm (or proceed if they said "just do it"). Then write every piece in full — no "you could write a post about X" stubs. While generating:
- Rewrite hooks natively per platform; never reuse the seed's opening line verbatim across pieces.
- Keep facts and claims within what the seed content supports — atomizing must not amplify a hedge into a certainty or invent numbers.
- Apply the one-CTA rule per piece, CTAs set by the goal from intake.
4. Sequence the distribution
Order the pieces for posting: lead with the strongest hook on the priority platform, don't post two derivatives of the same idea on the same platform back-to-back, thread/carousel mid-pack when early pieces have tested the angle. Present as a numbered posting order (the user maps it to their own dates/scheduler — no calendar prescriptions).
Required output format
# Atomization Pack — [seed content title]
## Idea inventory
| # | Idea (one line) | Type | Powers |
| 1 | ... | framework | Carousel, LinkedIn #2 |
(5–10 rows)
## Pack plan
Formats chosen: [list] — [one-line rationale tied to platforms/goal]
Skipped: [notable exclusions + why]
## Pieces
### LinkedIn Post 1 — [idea #]
[Full post text, ready to paste]
CTA: [the one ask]
### X Thread — [idea #s]
1/ [tweet]
2/ [tweet]
...
(char count noted per tweet)
### [Every other piece, fully written, grouped by platform]
## Posting order
1. [Piece] — [platform] — [why first]
2. ...
(strongest hook first; same-idea derivatives spaced apart)
## Leftovers
[Ideas from the inventory not yet used — future seed material]Quality bar
Check every piece before delivering:
- Every piece is fully written and paste-ready — zero outlines-as-deliverables (except formats that are outlines: carousel, deck, webinar)
- No two pieces share an opening line; no piece is a copy-paste excerpt of the seed
- Each piece within its format's length spec (count characters for tweets: ≤110 single, ≤250 per thread tweet)
- Exactly one CTA per piece, aligned with the intake goal
- Every claim/number traceable to the seed content or the user — nothing invented in translation
- Voice consistent across all pieces — read three at random; they must sound like the same person
- 5–7 formats, not more; each chosen format justified by a platform the user actually uses
- Posting order spaces out same-idea derivatives
Integration
skills/marketing/seo-content— primary source of seed articles; atomize each published article as a standard follow-up.skills/marketing/brand-voice— supplies the voice guide every piece must pass.skills/marketing/tweet-writer— takes over when the X pieces deserve dedicated niche research and hook testing; hand it the idea inventory.skills/marketing/newsletter— consumes the newsletter item; strong issues flow back here as seeds.skills/marketing/email-sequences— value emails can be atomized into social pieces and vice versa.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

