agentleFS
priyanshu-aryapriyanshu-arya/LinkedIn_Automation/.claude/skills/write-draft-x/SKILL.md

write-draft-x

Skill0 starsChanged 58 days agoMITView on GitHub

Use when the user asks to draft an X (Twitter) post or thread from a selected idea, run the X Writer, or explicitly invokes /write-draft-x. Converts an Idea Note claimed for the x platform into a Draft Note in Drafts/ (platform:x), grounded in its linked research and following Content-Learnings/voice-guide-x.md. Decides single-post vs. thread from the idea's depth. Does not assign posting days and does not handle approval or scheduling.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 2 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read the Idea Note and every research note in its sources[] — same verify-or-drop discipline as /write-draft: nothing in the post that …

    line 24
  2. Very specific:

    Read Content-Learnings/voice-guide-x.md in full, fresh each run.

    line 30
  3. Very specific:

    If this was the idea's only remaining unclaimed platform, flip status: selected → drafted.

    line 74
  4. Distinctive:

    Never invent a fact, statistic, or quote not present in the linked research notes.

    line 85
  5. Very specific:

    Never draft an idea that doesn't have an x entry in platform_schedule (run /plan-week-x first).

    line 87
  6. Very specific:

    Never silently copy another platform's draft text — this idea's X version is written fresh from the research, in voice-guide-x.md's voice, …

    line 89
  7. Distinctive:

    Never mark anything approved, scheduled, or published.

    line 92

≈ “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

Write Draft X (X Writer — X pipeline)

X's sibling to /write-draft. Converts one or more Idea Notes with an x entry in platform_schedule into Draft Notes (platform: x), written to Drafts/ using _Templates/Draft-Note.md. Assumes /plan-week-x has already assigned a day.

Arguments

  • idea-id (optional) — draft one specific idea for X. If omitted, draft every idea with an x entry in platform_schedule that doesn't already have an X draft linked to it (an idea can have both a LinkedIn draft and an X draft — they're independent notes, never copies of each other).

Process (per idea)

1. Load context

Read the Idea Note and every research note in its sources[] — same verify-or-drop discipline as /write-draft: nothing in the post that isn't already in the linked research's Summary/Key Findings.

2. Read the X voice guide and playbook

Read Content-Learnings/voice-guide-x.md in full, fresh each run. Also check Content-Learnings/playbook-x.md for an evidenced rule (real X post ids attached) about length, hook style, or single-vs-thread performance for this category — prefer it over the voice guide's generic defaults where one exists.

3. Decide single post vs. thread

Judge from the idea's actual depth, not a coin flip: if the point lands in one tight statement, write a single post. If it genuinely needs several connected beats (a numbered breakdown, a before/after, a multi-step argument), write a thread. Don't pad a single-post idea into a thread to look more substantial, and don't cram a multi-beat idea into one over-stuffed post.

4. Draft the post

  • Single post: hook-as-the-whole-post, tight, well under the character ceiling in voice-guide-x.md. Write the full text into ## Post Text and leave thread: [] empty.
  • Thread: first tweet is the hook (must stand alone if someone only sees tweet 1), each following tweet one beat, closing tweet lands the point without just recapping. Write the first tweet into ## Post Text (for at-a-glance consistency with single posts) and the full ordered sequence, first tweet included, into the thread: [] frontmatter list.
  • Same opinion/fact framing rule as /write-draft: a subjective take reads as a take ("I think," "my read"), not blended into declarative fact-voice.

5. Hashtags

0–2, only if genuinely useful for discovery, per voice-guide-x.md — don't default to LinkedIn's 3–5 convention here.

6. Write the Draft Note

Copy _Templates/Draft-Note.md into Drafts/ as YYYY-MM-DD--kebab-slug.md using the idea's platform_schedule entry for x (not target_date, which is LinkedIn's). Fill in: idea_id, platform: x, category, format, hook_style, hashtags, thread (per step 4), sources (inherited from the idea), status: draft, history: [{action: created, date: today, note: ...}]. Leave viral_score at 0 and visual_ids empty.

7. Update the idea

If this was the idea's only remaining unclaimed platform, flip status: selected → drafted. If other platforms in its platforms list haven't drafted yet, leave status: selected so they can still pick it up.

Report back

Show the full draft (all tweets in order if a thread), hashtags, per-tweet character counts, and which research sources backed it.

Hard rules

  • Never invent a fact, statistic, or quote not present in the linked research notes.
  • Never draft an idea that doesn't have an x entry in platform_schedule (run /plan-week-x first).
  • Never silently copy another platform's draft text — this idea's X version is written fresh from the research, in voice-guide-x.md's voice, even if a LinkedIn draft for the same idea already exists.
  • Never mark anything approved, scheduled, or published.

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.