agentleFS
busyeugenebusyeugene/content-marketing-skills/repurpose-threads-x/SKILL.md

repurpose-threads-x

Skill1 starsChanged 7 months agoMITView on GitHub

Use when the user wants to repurpose an article, talk, or long-form content into a Threads or X (Twitter) thread. Extracts the strongest points, sequences them into tweet-sized atoms, drafts multiple variants, and optionally publishes.

  • Reads credentials

The file

Show its rules and text

We split this skill into the rules it gives an agent

15 rules. About 0 are common, 12 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈12

    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
  1. Distinctive:

    No required keys. Optional (only for publishing):

    line 15
  2. Very specific:

    Use AskUserQuestion for platform, length, and angle.

    line 28
  3. Very specific:

    Read editorial-guidelines.md if present.

    line 34
  4. Distinctive:

    Read the source file or scrape the URL.

    line 35
  5. Distinctive:

    Never "Here's a thread on X" or "1/".

    line 54
  6. Distinctive:

    Use line breaks inside a tweet sparingly.

    line 63
  7. Distinctive:

    Avoid "next up", "continuing", "and another thing".

    line 64
  8. Distinctive:

    Default max 280 chars per tweet.

    line 77
  9. Distinctive:

    No hashtag spam (0–2 max, only if they read naturally).

    line 79
  10. Distinctive:

    Avoid emojis unless the brand voice explicitly allows them.

    line 80
  11. Distinctive:

    No banned words from editorial guidelines.

    line 91
  12. Distinctive:

    No tweet is filler ("Now let's talk about…", "Another point is…").

    line 93
  13. Very specific:

    Only if the user explicitly asks AND X credentials are set.

    line 143
  14. Distinctive:

    Source unavailable: ask for a paste.

    line 153
  15. Distinctive:

    No editorial guidelines: use shared voice reference.

    line 155

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Repurpose → Threads / X Thread

Turn a long-form source into a tight thread for Threads or X. Produces multiple variants with different framings and saves each as a local file. Publishing is optional and delegated to the social-post skill's publish path if the user asks.

Setup

No required keys. Optional (only for publishing):

  • X: X_API_KEY, X_API_SECRET, X_ACCESS_TOKEN, X_ACCESS_TOKEN_SECRET
  • Threads: no official public API at time of writing. Threads posts are draft-only from this skill; the user posts manually.

Inputs

  1. Source — path to a file or a URL.
  2. Platform — x, threads, or both. Different character limits and cultural norms per platform.
  3. Thread length — short (5–7 tweets), medium (8–12), long (13–20). Default medium.
  4. Angle — one or more of: listicle, contrarian, framework, story, data. Default: ask.
  5. Number of variants — default 2 per platform.
  6. Publish? — default no.

Use AskUserQuestion for platform, length, and angle.

Process

1. Load voice and source

  • Read editorial-guidelines.md if present.
  • Read the source file or scrape the URL.

2. Extract the atoms

Pull from the source:

  • The single strongest claim
  • 10–15 supporting atoms (one fact, one example, one data point each)
  • 2–3 numbers that can anchor tweets
  • 1–2 potential "quote-tweet bait" lines — short, specific, opinionated

Tag each atom with its best use: hook, body, stat, closer.

3. Draft the thread

For each requested variant, sequence the atoms into a thread with this structure:

T1 — hook (most important tweet):

  • Stand-alone line that makes sense without context.
  • States a specific claim, number, or contrarian take.
  • Never "Here's a thread on X" or "1/".
  • ≤200 chars even on X Premium; mobile readers don't scroll a 280-char T1.

T2 — why it matters:

  • The stakes for the reader in one sentence.

T3 through T(N–2) — body:

  • One atom per tweet. No run-on sentences across tweets.
  • Each tweet can be quoted standalone.
  • Use line breaks inside a tweet sparingly.
  • Avoid "next up", "continuing", "and another thing".

T(N–1) — the punchline or turn:

  • The one tweet the reader should remember. Often the most contrarian line.

T(N) — closer + CTA:

  • One-line summary of the thread.
  • One CTA only: follow, reply to a specific question, DM a keyword, or click a link. Never two.
  • Optional PS tweet with the link only, so the CTA tweet stays clean.

4. Platform-specific constraints

X:

  • Default max 280 chars per tweet. Note long-tweet mode (up to 25,000 chars for X Premium subscribers) in the file header if the user has Premium and asked for it.
  • Posting requires the Basic tier ($100/month) or higher. The Free tier returns 403 on POST /2/tweets. If the user only has Free, draft locally and skip publish.
  • No hashtag spam (0–2 max, only if they read naturally).
  • Avoid emojis unless the brand voice explicitly allows them.
  • No "1/" numbering unless the user asks — it reduces retention.

Threads:

  • 500 chars per post, so atoms can breathe more. Use the extra room for specificity, not fluff.
  • Threads audience tolerates more personal/narrative framing than X.
  • Link previews render better on Threads — the final post should include the link as its own line.

5. Voice checks

Before writing:

  • No banned words from editorial guidelines.
  • First tweet passes the "would a tired scroller stop?" test.
  • No tweet is filler ("Now let's talk about…", "Another point is…").
  • Numbers are cited or qualified ("based on our Q3 data", "according to {source}").

6. Write the files

One file per platform per variant under threads/ (for Threads) or x/ (for X).

Path: threads/<slug>-v{n}.md or x/<slug>-v{n}.md.

# {platform} thread — {slug} v{n}

**Source:** {path or URL}
**Angle:** {angle}
**Length:** {n} tweets
**Generated:** {YYYY-MM-DD}

---

**T1** (hook) — {char count}

{tweet body}

---

**T2** — {char count}

{tweet body}

---

{…}

---

**T{N}** (closer) — {char count}

{tweet body}

---

## Rationale

- Hook chosen: {which atom, why}
- Punchline: {T(N-1) atom}
- CTA: {action}

7. Optional publish

Only if the user explicitly asks AND X credentials are set. Delegate to the social-post skill's publish path (X platform, thread format) with the chosen variant. Threads has no public API — always draft-only.

Confirm once more with the user before the first publish call.

8. Print summary

One block: files written, tweet counts per variant, recommended variant with one-line reason, publish status.

Fallbacks

  • Source unavailable: ask for a paste.
  • Atoms fewer than requested length: suggest shorter thread length rather than padding.
  • No editorial guidelines: use shared voice reference.

Verification

  1. One file per requested variant under the right platform directory.
  2. Each tweet shown with its char count.
  3. T1 stands alone as a quotable line.
  4. Exactly one CTA in the final tweet (optional PS for link).
  5. Char counts all within platform limits.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.