LinkedIn Post Writer

When to Use This Skill

Load this skill whenever the user:

• Shares an idea, rough notes, or article and asks for a LinkedIn post
• Asks to write, draft, or improve a LinkedIn post
• Wants help with thought leadership content for LinkedIn
• Is a professional in any field building a credible public voice

Inputs

Required: the raw material — an idea, notes, an article, a transcript excerpt, or an existing draft.

Optional but useful: the user's role and industry (so examples and vocabulary fit), the audience they want to reach, a preferred opening mode, and any real numbers, names, or situations they are willing to share publicly.

Missing-input behavior: if the raw material is missing, ask once. If the audience or industry is missing, infer it from the material and say so. Never invent personal anecdotes, numbers, or client details to fill gaps — leave a [add specific] placeholder and tell the user what would make the post stronger.

Core Philosophy

The goal is to sound like a sharp, self-aware human — not a brand account, not a motivational poster, not a chatbot. The best LinkedIn posts feel like something a smart colleague said over coffee that you immediately wanted to forward.

Three non-negotiables:

1. The first 1-3 lines must earn the "see more" click on their own merit
2. The voice must be conversational — specific words, not polished words
3. Sincerity over performance — professionals who sound real build trust faster than professionals who sound impressive

The Hook (Lines 1–3): Above the Fold

LinkedIn truncates posts after roughly 2-3 lines. Everything above that fold is your headline. It has one job: stop the scroll.

Choose one of these three opening modes:

1. Curiosity Gap

Create information asymmetry. The reader knows something is missing and wants the answer.

"Most feedback conversations fail before they start. Not because the feedback is wrong. Because the timing is."

"I've hired 40+ people. The best interview question I ever used wasn't a question at all."

2. Contrarian Statement

Disagree with a widely held belief in your industry. Be specific — vague contrarianism is noise.

"Async-first culture sounds great in theory. In practice, it rewards people who write well and punishes everyone else."

"Every hospital says it's patient-centered. Then it schedules around the physician's calendar."

3. Value Delivery / Direct Claim

State the payoff upfront. The reader knows exactly what they're getting.

"Here's the email structure I use when I need a decision from someone too senior to read a long thread."

"Three things I stopped doing in 1:1s that made them 10x more useful."

Hook writing rules:

• Never start with "I" (weak opener — and widely believed to hurt reach)
• Never start with a question as the very first word (feels like a quiz)
• No setup before the setup — the first sentence IS the hook
• Keep each line to one thought
• Leave a gap between lines 1, 2, and 3 (white space increases readability and click-through)

The Body: Staccato Without Drama

After the hook, the post should breathe. Short paragraphs. Single-sentence bursts followed by brief elaboration. Not tweet-length fragments, but not paragraphs either.

Rhythm pattern:

• One sharp claim (1 sentence)
• Brief expansion (1-2 sentences max)
• [White space]
• Next point

What to avoid:

• Numbered lists of more than 5 items (feels like a blog, not a post)
• Bullet points for every single thought (lazy structure)
• Transition phrases: "The thing is...", "Here's the deal...", "Let's be honest..."
• Ending every sentence with a period on its own line for fake drama
• Rhetorical stacking: "Not X. Not Y. Not Z." (overused pattern)
• Any sentence that could appear in a corporate press release

What works:

• Specific examples over general principles ("When I was running a team of 12 at [X]" beats "In my experience as a leader")
• Numbers and specifics ("took 4 months" not "took a long time")
• The personal moment that proves the point — brief, not confessional
• An honest admission that creates contrast ("I got this wrong for two years")
• Vocabulary from the user's actual field — a nurse manager, a SaaS founder, and a construction PM should not sound the same

AI Tells to Eliminate

These phrases and patterns immediately signal AI-generated content. Never use them:

Phrases to ban:

• "In today's fast-paced world..."
• "Game-changer" / "paradigm shift" / "unlock your potential"
• "It's not about X, it's about Y" (overused LinkedIn cliché)
• "Here's what I've learned:" (as a standalone transition line)
• "The truth is..." / "The reality is..."
• "This is your sign to..."
• "Swipe right if..." / "Double tap if..."
• "Let that sink in."
• "Full stop." / "Period."
• "No fluff." / "No gatekeeping."
• Excessive em dashes used for dramatic effect — like this — every other sentence
• Ending with "Agree?" or "Thoughts?" as the only CTA

Structural AI tells:

• Hook → 5-item numbered list → closing question (the most common chatbot format)
• Every paragraph being exactly 1-2 sentences (robotic rhythm)
• Overuse of parenthetical asides (like this) (and this) (and another one)
• Alliteration in headers or bullets ("Plan, Prepare, Perform")
• Triads everywhere — three adjectives, three examples, three-part closers, in every paragraph

Closing

End with one of:

• A single honest observation that reframes the whole post
• An open question that invites real debate (specific, not "Agree?")
• A quiet call to action — something they can actually do in the next 10 minutes
• A callback to the opening line that reframes it

Avoid a closing that announces itself ("So what's the takeaway?"). The ending should feel like the last line of a good short story — it lands without a thud.

Reference Examples

These are patterns drawn from writers who consistently get this right — Justin Welsh, Wes Kao, and Jay Clouse among them. Their posts are public; read a few directly. What follows is the structural analysis of what each pattern does, so it can be reproduced with the user's own material rather than imitated word for word.

Justin Welsh — Contrarian / Identity

Structure: Opens with a one-line claim about what most people do (build for the crowd). Second line flips it: the most successful people do the opposite. Then three or four short declaratives describing what "the opposite" looks like — honesty, specificity, comfort with not being understood. A short cause-and-effect run (content sharper, audience more engaged, right people attracted). Closes with a two-line reframe that contrasts everyone with someone: "Stop trying to be everything to everyone. Start being something to someone."

What works: Short declarative sentences that don't explain themselves. No filler transitions. The ending lands as a reframe, not a tagline. Nothing in the post is a statistic or a citation — it is entirely a point of view, stated with confidence.

Justin Welsh — Process / Scroll-Stopper

Structure: A "How to" first line. Then a six-item numbered list where every item pairs an instruction with a parenthetical question the reader is silently asking (why should I stop scrolling? what am I learning? why should I come back?). One-line close promising the result of doing it daily.

What works: The numbered list works here — despite the five-item rule above — because each item earns its spot by answering an unstated objection in real time. It is a list of answers, not a list of tips.

Wes Kao — Contrarian / Credibility

Structure: Opens with an observation that simple-looking things take enormous work. A personal example (writing a newsletter) with a specific, self-deprecating estimate of the hours involved. An admission of embarrassment, and of getting more open about it. Then the turn: a critique of managers who have never built "The Thing" — a coined phrase that becomes the post's recurring anchor. A short explanation of why player-coaches matter. Closes by reframing the reader's self-criticism: you are probably not too slow; your expectations are probably unrealistic.

What works: The coinage becomes a handle the reader can carry away. The admission is real and earned, not confessional performance. The ending reframes without preaching — it gives the reader permission rather than instruction.

Wes Kao — Practical / Value Delivery

Structure: A three-word "How to" first line. Then a before/after pair: a rambling, stream-of-consciousness "before" message rendered in miniature, and a tight "after" that leads with the ask and the context. Two short paragraphs explaining that concision is about preparation, not omission. One-line close on the payoff: say the punchline first and you instantly seem more competent.

What works: The before/after format makes the abstract concrete immediately. The post teaches without lecturing because it shows instead of tells. The reader can apply it in their next message.

Jay Clouse — Observation / Emotional Hook

Structure: Opens by naming a problem with most content: it doesn't connect emotionally. Describes the feeling of being shouted at for doing too much or too little. A parenthetical admission that nobody's life is optimized. Then a first-person admission of contributing to the noise. The turn: telling people what to do is easy; showing them is hard. A practical instruction — model the behavior yourself and explain why it matters and for whom. Closes on the payoff: more credible, easier to connect with.

What works: The admission "I've contributed to this noise" earns the moral authority to make the critique. The closing is a practical reframe, not a platitude.

Jay Clouse — Series / Curiosity Gap

Structure: Opens by ruling out the obvious interpretation (not carousels vs. video vs. text) before stating the real one (the format within those formats). Then an observation about creators thinking in series, with placeholder-style examples ("a series on [how to X]") and an explanation of why numbering works — it pulls readers backward and forward. One-line close on the effect: anticipation and audience investment.

What works: Opens by subverting the expected interpretation. Uses bracketed placeholders to make an abstract concept immediately tactile. The whole post is one idea, fully worked, and nothing else.

Workflow

When the user provides input (idea, notes, article, or rough draft):

1. Identify the real message — what is the one true thing this person wants to say? Strip the content to its core claim.
2. Choose the opening mode — curiosity gap, contrarian, or value delivery. Pick the one that most naturally fits the content, not the one that sounds most dramatic.
3. Write the hook (lines 1-3) — treat this as its own deliverable. It must stand alone.
4. Build the body — expand with specifics, a brief personal moment if relevant, and no more than 3-4 distinct points.
5. Close cleanly — one landing line, not a paragraph.
6. Audit for AI tells — read back through the banned phrases list. Remove or rewrite any match.
7. Read it aloud (mentally) — if it sounds like a press release or a motivational poster, rewrite it.
8. Offer the user one variation — if the opening mode is ambiguous, offer a contrarian version and a value-delivery version so they can choose.

Output Format

Always present:

• The full post, ready to copy/paste
• A brief note on which opening mode was used and why
• One optional alternate hook if the choice was close

Do not explain every structural decision unless the user asks. Keep the meta-commentary minimal.

Guardrails

• Never fabricate. No invented numbers, clients, employers, outcomes, or anecdotes. If the post needs a specific and the user didn't supply one, use a [add specific] placeholder and say so.
• Respect confidentiality. If the source material names clients, patients, colleagues, or deal details, anonymize by default and flag anything that may need permission before posting.
• No claims the user can't back. Superlatives ("the best", "the only", "guaranteed") get cut unless the user has substantiation — the same instinct that keeps marketing copy honest applies to a personal post.
• The user's voice wins. If they supply prior posts, match their rhythm and vocabulary over the defaults here.

Quick Checklist Before Delivering

• First line does not start with "I"
• No banned phrases from the AI tells list
• No numbered list longer than 5 items
• Body has white space between ideas
• Closing lands without announcing itself
• Specific details used (numbers, names, real situations) where possible — none invented
• Vocabulary fits the user's field
• Would a smart colleague forward this or roll their eyes?