linkedin-post
Skill1 starsChanged 23 days agoMITView on GitHub
Write a LinkedIn post for one of your businesses, in that business's topics and tone, and show exactly what the feed shows before "see more" on mobile and desktop. Use it whenever someone says "write a LinkedIn post", "post about this", "turn this into a LinkedIn post", "draft Tuesday's post", "I want to share that we hit X", or pastes a win, a lesson or a story and wants it on LinkedIn. Works across several businesses: it picks the active one from founder-context, one business per post. Gives hook options scored by a local tool, one full draft cleaned by humanize, a fold preview, and a copy-ready block. It never posts for you.
44 other projects use a version of this.
Adopted by
Most recently taken up by
- Lindadao92/needle-content-skillstheir own version, first seen Nov 16, 2026
- runehub-ai/rune-hubtheir own version, first seen Nov 15, 2026
- dmccreary/claude-skillstheir own version, first seen Nov 9, 2026
- dmccreary/ibook-skillstheir own version, first seen Nov 8, 2026
- kristian-presso/claude-marketing-skillstheir own version, first seen Nov 8, 2026
- olivMertens/copilot-skillstheir own version, first seen Nov 8, 2026
Used with
Projects with it also use youtube-upload, proposal-writer and 3 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
BlackLotus1974Skillyoutube-upload
Upload videos to Atchalta's YouTube channel (office@atchalta.com). Use when the user asks to upload a video to YouTube, list YouTube videos, manage playlists, or update video metadata. Trigger phrases: 'upload to youtube', 'youtube video', 'publish video', 'post to youtube', 'list my videos', 'youtube playlists'.
Also in 3 projects that use this
ncklrsSkillproposal-writer
Expert sales proposal and pricing presentation strategist. Use when writing proposals, executive summaries, ROI business cases, pricing presentations, SOWs (Statement of Work), RFP responses, or competitive positioning documents. Covers proposal structure, terms positioning, design formatting, follow-up strategy, and win-rate optimization.
Also in 3 projects that use this
Microsoft AzureSkilllint
Run all linting and spell checking for TypeScript and Rust code. Run this before committing to catch style issues, clippy warnings, and misspellings.
Also in 3 projects that use this
OpenAISkillbrainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
Also in 3 projects that use this
GoogleSkillgrill-me
A relentless interview to sharpen a plan or design.
Also in 3 projects that use this
Other versions44 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 44
- langchain-ai/langgraph-101★ 708same name
- hamzafarooq/claude-code-starter★ 145same name
- dmccreary/ibook-skills★ 112same name
- dmccreary/claude-skills★ 110same name
- kevinbadi/ai-os-skills★ 69same name
- publora/skills★ 68same name
- az9713/ai-co-writing-claude-skills★ 61same name
- JamsusMaximus/growth-agents★ 21same name
- langchain-ai/langgraph-101-ts★ 19same name
- duncan-buildroom/profit-room-skills★ 18same name
34 more not listed
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 repository15 other files it gives its agents.
- founder-contextSkill
- humanizeSkill
- linkedin-auditSkill
- linkedin-carouselSkill
- linkedin-commentSkill
- linkedin-dmSkill
- linkedin-hooksSkill
- linkedin-inboxSkill
- linkedin-planSkill
- linkedin-profileSkill
- linkedin-replySkill
- linkedin-repurposeSkill
- linkedin-viralSkill
- talking-head-cardsSkill
- transcribeSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
19 rules. 0 are common, 15 are distinctive, and 4 are very specific to this project.
0
Common
Most of the 38 edited versions we read keep these.
15
Distinctive
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 31
If the user names a business ("for Frameleaf"), use it.
- Distinctive: line 35
If none exist, ask once: "Which business is this for?
- Very specific: line 44
If it has plan.md, check whether this post fills a planned slot and use that slot's angle.
- Very specific: line 45
If it has posts.csv, read the last 10 rows so you do not repeat a recent topic or hook pattern.
- Distinctive: line 51
If the idea is thin ("post about AI"), ask one batched question for the specific moment, the number, and who it was for.
- Distinctive: line 53
If the idea is off the business's topics, say so and ask whether it belongs to another business or should be skipped.
- Distinctive: line 55
If the idea mixes two businesses, offer two posts.
- Distinctive: line 67
If the tool is missing, judge by hand on the same things: fits before the fold, concrete, has tension, a person in it, plain, no bait.
- Distinctive: line 75
Write the full post on the chosen hook.
- Very specific: line 89
Use only proof from the business file or the user's own words.
- Distinctive: line 92
Do not invent framing either: no claims about what other people feel or do ("office managers always..."), no made-up feelings for the …
- Distinctive: line 100
No link in the body. Put it in the first comment and say so in the receipt.
- Distinctive: line 102
No emoji unless the business file says the founder uses them.
- Very specific: line 106
Run the draft through the humanize skill before showing it.
- Distinctive: line 123
Run it on the final text. If you change the draft after the check, run it again.
- Distinctive: line 152
Never log a post the user has not confirmed.
- Distinctive: line 171
Never post, schedule, log in, or use any tool that touches LinkedIn for the user.
- Distinctive: line 173
Never write for two businesses in one post.
- Distinctive: line 174
Never make up a fact, a number, a customer, or a quote.
“Common” is measured: kept by at least 6 in 10 of the 38 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Based on li-post by Jake Schincariol (MIT). Changed by Aya Taleb: one business per post from founder-context, a measured "see more" preview, per-business logs, humanize, scored hooks.
linkedin-post
Turns one real thing that happened into a LinkedIn post for one business. The user posts it. This skill only writes.
Tool paths: <this skill> is the folder this SKILL.md is in (the "base directory" you
were given). Sibling skills sit next to it, so <this skill>/../linkedin-hooks/ is the
hooks skill. Run tools with full paths built from that folder, not from the current
working directory, or they will look missing.
1. Find the business
The founder may run several ventures. Every post belongs to exactly one, because a reader who follows you for office lunches does not want your app's beta notes.
- If the user names a business ("for Frameleaf"), use it.
- Else, if the project has
founder/context.md, use that. - Else read the slug in
~/.founder/activeand open~/.founder/businesses/<slug>.md. - If none exist, ask once: "Which business is this for? Give me the name, who it serves, 3 tone words, words to avoid, and any proof you can use." Offer the founder-context skill to save it. Never invent this file's contents.
Say which business you are writing for in your first line, even when you are refusing part of the request, so a wrong guess is caught early. From the file, take: topics, tone (3 words and words to avoid), proof, off-limits.
The business folder is ~/.founder/social/linkedin/<slug>/. Create it if missing.
If it has plan.md, check whether this post fills a planned slot and use that slot's angle.
If it has posts.csv, read the last 10 rows so you do not repeat a recent topic or hook pattern.
2. Check the idea before writing
A post needs one specific true thing: what happened, to whom, and what it cost or gave back.
- If the idea is thin ("post about AI"), ask one batched question for the specific moment, the number, and who it was for. Do not pad a thin idea into a long post.
- If the idea is off the business's topics, say so and ask whether it belongs to another business or should be skipped. If it touches an off-limits item, stop and say which one.
- If the idea mixes two businesses, offer two posts.
3. Hooks: write five, score them, show three
Line 1 is the only line most people read. Write 5 different opening lines, each a different approach (a number, a before and after, a mistake, a scene, a plain claim, a question only this founder could ask). Then score them with the linkedin-hooks tool:
python3 <this skill>/../linkedin-hooks/scripts/hookscore.py hooks.txt --topics "<topics>" --avoid "<words to avoid>"If the tool is missing, judge by hand on the same things: fits before the fold, concrete, has tension, a person in it, plain, no bait. Show the top 3 with their scores and say which you would ship in one sentence. If you reword a hook, score the new wording again before you use it, and show that score. The score ranks options; your judgment and the user's voice win a tie.
4. Draft
Write the full post on the chosen hook.
Line 1 the hook, alone on its line, under about 140 characters
Line 2 pays off line 1 (what happened), not setup for line 3
Body short paragraphs, 1 to 3 lines, a blank line between them
Turn one line that says what it means
Close one specific question or one instruction, never both- 600 to 1,300 characters for most posts. The hard limit is 3,000.
- Build the post on what the user told you. Add a fact from the business file only when the post needs it to make sense, and at most one or two. A post stuffed with every proof point reads like an ad.
- Use only proof from the business file or the user's own words. A missing number becomes
{{your number}}, flagged in the receipt. Never invent a client, a quote, a result or an example person to show a point. Proof is cleared for use; only the off-limits line is off limits. - Do not invent framing either: no claims about what other people feel or do ("office managers always..."), no made-up feelings for the founder, no thanks to people the user did not mention, and no "we never pitched them" or "they got in touch" unless the user said so. Proof is a fact, not a promise: "average drop 12:15" does not become "see you at 12:15". A number you work out yourself is new too: if 6 offices are on plans and one just signed, say "our 7th office", not "7 offices on weekly plans". When in doubt, cut the line. A short true post beats a longer one the founder has to correct.
- The business's tone words set the voice. Its words to avoid never appear.
- No link in the body. Put it in the first comment and say so in the receipt.
- Up to 3 hashtags at the end, only real ones people follow. None is fine.
- No emoji unless the business file says the founder uses them.
5. Clean it
Run the draft through the humanize skill before showing it. If humanize is not
installed, follow references/writing-check.md by hand and say you did. Report what
changed in one line. A post that reads as machine-written costs the founder trust.
6. Show the fold
Save the draft to ~/.founder/social/linkedin/<slug>/drafts/<YYYY-MM-DD>-<short-topic>.md
through the log tool, so it lands in the right folder and never overwrites by accident. Write it in
the project first, then copy it in and delete the project copy:
python3 <this skill>/scripts/postlog.py save --from <project draft> --to <that path>.
Then run:
python3 <this skill>/scripts/fold.py <draft file> # text post
python3 <this skill>/scripts/fold.py <draft file> --media # with an image, video or PDFRun it on the final text. If you change the draft after the check, run it again. Copy its
two previews into your reply, word for word, as two labeled blocks (MOBILE and
DESKTOP). Tool output is not visible to the user, so a preview left in the tool result
counts as not shown. It shows the part the feed shows before "see more", on mobile and on
desktop. These cutoffs are observed, not published by LinkedIn, and were checked on
2026-10-23 (sources in references/limits.md). If the hook is cut off on mobile, or the
desktop view ends mid-sentence in a way that reads badly, fix the lines and run it again. If the tool cannot run, count by hand: about 3 lines or
140 characters on mobile, about 5 lines or 210 on desktop, a blank line counts as a line.
7. Hand it over
Print the post in one fenced block, exactly as it should be pasted. Then:
POST READY (business: <name>)
hook: <pattern>, score <n>
length: <n> characters
fold: mobile shows "<...>" | desktop shows "<...>"
cleaned: <one line from humanize or the writing check>
fill in: <each {{placeholder}}, or "nothing">
first comment: <the link, if any>
when: <slot from plan.md, or "your call">Then say: "Post it yourself on LinkedIn. Tell me when it is live and I will log it." Add one line naming the next skill that fits (see the end of this file).
8. Log it when the user says it is posted
Never log a post the user has not confirmed. On "posted" or "done", append one row:
python3 <this skill>/scripts/postlog.py add --log ~/.founder/social/linkedin/<slug>/posts.csv \
--date <YYYY-MM-DD> --type <proof|opinion|teach|story|offer> --topic "<topic>" \
--pattern "<hook pattern>" --first-line "<line 1>" --chars <n>The number columns stay blank. The user fills them in later from LinkedIn analytics, and linkedin-audit reads them. linkedin-plan reads the log to rotate topics.
When the user asks you to post, log in, or schedule
Say no in the first lines of your answer, in plain words: you do not log in, post or schedule, and the user posts it themselves. If they shared a password, do not repeat it, and suggest they change it. Then carry on with the writing part of the request.
Never
- Never post, schedule, log in, or use any tool that touches LinkedIn for the user. LinkedIn's User Agreement bans automation, and an automated account can be restricted.
- Never write for two businesses in one post.
- Never make up a fact, a number, a customer, or a quote.
What runs next
- linkedin-hooks, for more hook options on the same idea.
- linkedin-carousel, if the idea is a how-to that would work as slides.
- linkedin-reply, once comments come in.
- linkedin-plan, to see the next slot.
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.

