LinkedIn Growth8 min read
14 Claude Prompts for LinkedIn Posts You Can Copy and Paste
A copy-paste library of 14 Claude prompts for LinkedIn posts: hooks, story posts, contrarian takes, carousels, and rewrites that strip the AI tells.
By Animesh Kudake, Co-founder, Apna Project
The short version
- The best Claude prompts for LinkedIn posts are built for one job each: a hook, a story, a contrarian take, a carousel, or a rewrite that strips the AI tells.
- A prompt library beats one-off prompting because you stop re-explaining yourself and start iterating on a draft that already carries your context.
- Paste three or four of your own past posts as a voice sample before Claude writes anything. This single habit does more for tone than any wording trick.
- Save your instructions and voice sample in a Claude Project so every new post starts from your context instead of a blank chat.
- Prompts get you a fast draft. The editing, the fact-checking, and the day-to-day consistency are still yours to own.
Prompts that turn what you know into posts
Your best posts already live in your head as a lesson, a memory, an argument, or a number.
Turn this lesson into a LinkedIn story post: "[the lesson in one sentence]". Structure it as a short scene, then the turn, then the lesson, then one line the reader can use today. Under 150 words, first person, plain language. The lesson should feel earned by the story, not tacked on.
Use it when you have a takeaway but no story around it. Adapt by giving the real who, when, and what went wrong so the scene stays specific.
Here are rough notes from something that happened to me: [3 to 5 messy bullet points]. Draft a LinkedIn post from these in my voice. Keep the mess that makes it real: the specific numbers, the awkward moment, the thing I got wrong. Do not add fake drama or a neat moral. End on the honest takeaway.
Use it when a real experience should not read like a fable. Adapt by naming the emotional core so Claude does not smooth it away.
The common advice in my field is: "[widely repeated belief]". I disagree. Help me write a contrarian LinkedIn post arguing the other side. Steelman the popular view in one line first so I do not look lazy, then make my case with a concrete example. Confident, not rude, no strawmen. End with a question that invites disagreement.
Use it when you hold a real, defensible unpopular view. Adapt by feeding Claude your evidence so the argument stands on something.
Here is a finding from my own work: [the number or result, and how you got it]. Write a LinkedIn post that leads with the number, explains what surprised me, and gives the reader one thing to do differently. Do not invent extra statistics. If a claim needs a source I have not given you, flag it instead of filling it in.
Use it when you have a real result to share. Adapt by keeping the no-invented-statistics line.
Prompts for carousels and repurposing
One idea becomes a document post; one article becomes a week of posts.
I want a LinkedIn carousel about [topic] for [audience]. Give me a slide-by-slide outline: slide 1 is the hook, slides 2 to 7 each make one point in under 15 words, and slide 8 is a soft call to action. For every slide, write the on-slide text and a one-line note on the visual.
Use it when you want a document post but face empty slides. Adapt by setting the slide count and the one idea per slide.
Here is an article: [paste the text or key points]. Pull out the one idea most likely to resonate on LinkedIn and write a standalone post about it in my voice. Do not summarize the whole thing. Pick the sharpest angle and make it self-contained. Then suggest 2 more angles I could use for future posts.
Use it when you have long-form content and want several posts from it. Adapt by asking for the extra angles.
Prompts for engagement and replies
Reactions are cheap. Comments are the signal that drives LinkedIn engagement, and both prompts are built to earn them.
Write a short LinkedIn post that ends in a question my network will actually answer. Topic: [topic]. The question must be easy to answer from experience, specific enough to avoid one-word replies, and not something people feel tested by. Give me 3 versions with different questions, each under 80 words.
Use it when you want comments, not just likes. Adapt by aiming at a decision people have made, not an abstract opinion.
Someone commented this on my post: "[their comment]". My post was about [topic]. Draft 3 short replies in my voice that move the conversation forward: one that adds a detail, one that asks a genuine follow-up, and one that respectfully pushes back. Each under 30 words. No "thanks for sharing" filler.
Use it when a post takes off and you want replies that keep the thread alive. Adapt by answering the sharpest comments first, using our LinkedIn comment examples for the shape.
Prompts that make Claude sound less like AI
A model gives you a competent draft, plus the tells: even rhythm, hype words, lines that could sit under anyone's post. These three take them out.
Rewrite this post so it stops sounding like AI wrote it: [paste draft]. Remove hype words, vary the sentence length, cut any line that could sit under a different post unchanged, and delete filler transitions. Keep my contractions. Do not add new claims. Return only the rewrite.
Use it when a draft reads a little too smooth. Adapt by keeping a ban list of words you never say and pasting it in.
Rewrite this draft to match the voice sample I gave you earlier: [paste draft]. Match my average sentence length, my level of formality, and my vocabulary. If a line sounds more like you than me, flag it and suggest my version. The goal is that a colleague cannot tell a model was involved.
Use it when the draft is right but the voice is off. Adapt by keeping the setup voice sample in the same chat.
Act as a tough LinkedIn editor. Here is my draft: [paste]. Do not rewrite it yet. Tell me the weakest line, the line I should lead with instead, one thing that is still vague, and whether the ending earns a comment. Then give me a tightened version that is 20% shorter.
Use it when a post is fine but not sharp. Adapt by fixing the shorter length so Claude has to make real cuts.
What a prompt library can't do for you
Every prompt here gets you a draft, not a good post. What is true, what is worth saying, and what crosses a line stay with you. So does consistency: a model matches your voice in one chat and drifts in the next, which is why the voice sample and a saved Project matter more than any clever prompt.
Prompts also will not show up every day. Posting is half of LinkedIn; the rest is commenting where your buyers already read, a daily habit most people quit. For the mechanics, we wrote a full commenting strategy and an honest look at whether AI comments belong on LinkedIn.
That daily habit is what our own tool, Commenti, is built to carry: it watches the creators and keywords you pick and drafts comments in your tone, anchored to a specific line in each post, on your approval or on a schedule with caps and human-like pacing. We will be straight about the trade. LinkedIn's User Agreement prohibits third-party automation, ours included, so the caps and the approve mode reduce risk. Nothing erases it.
Prompts write the draft. Your voice, your judgment, and your consistency turn a folder of drafts into a following.
Frequently asked questions
What is the best Claude prompt for writing a LinkedIn post?
There is no single best prompt, because different posts need different jobs. Use a hook prompt for openers, a lesson-to-story prompt for narrative posts, and a rewrite prompt to strip robotic phrasing. The real gain comes from pairing any prompt with a voice sample of three or four of your own past posts so the draft sounds like you.
Which Claude model should I use to write LinkedIn posts?
Any current model at claude.ai will draft a post. Opus 4.8 is the most capable for nuance and voice matching, the Claude 5 models like Sonnet 5 and Fable 5 are strong all-rounders, and Haiku 4.5 is the fast one for quick idea lists. Your prompt and voice sample matter more than the model you pick.
How do I make Claude write posts that do not sound like AI?
Paste a few of your own posts as a voice sample, then ask Claude to match your rhythm and vocabulary. Ban stock phrases outright, keep sentence length varied, and run a final pass that removes hype words. Then edit by hand. The last bit of humanity in a post almost always comes from you, not the model.
Can I use Claude prompts to write LinkedIn comments too?
Yes, the same idea works for replies. A comment-reply prompt fed the original post and your angle produces a specific response instead of generic praise. For comments at real volume, a purpose-built tool like our own Commenti drafts in your tone, though LinkedIn's User Agreement prohibits third-party automation, which carries risk you should weigh.
Written by the team building Commenti on real LinkedIn growth data. Found an error? Tell us and we’ll fix it. Accuracy beats winning.

Co-founder, Apna Project
Co-founder at Apna Project, where he helps founders and agencies turn ideas into products. Writes here on LinkedIn growth and what actually earns attention in a comment section. Commenti is the LinkedIn comment automation tool by Ampliflow.
The best Claude prompts for LinkedIn posts are not clever one-liners. They are small, repeatable instructions built for one job each: writing a hook, turning a lesson into a story, arguing a contrarian point, or stripping the AI tells out of a draft. Below are 14 you can copy, paste, and adapt today.
Copy the shape, not the sentences; the
[brackets]mark what you swap in.Why a prompt library beats one-off prompting
One-off prompting means re-explaining who you are every time you open a chat. A saved library fixes three things at once.
Context. A prompt works better when Claude knows your topic, your audience, and the outcome you want. Bake that in and you stop starting from zero.
Your voice. Paste three or four of your own posts and ask the model to learn the rhythm before it writes a word. This does more for how a post sounds than any wording trick.
Iteration. When a draft comes out wrong, you fix the prompt, not just the post, and the next fifty posts inherit the fix.
At claude.ai any current model works: Opus 4.8 is the most capable for voice matching, the Claude 5 models like Sonnet 5 and Fable 5 are strong all-rounders, and Haiku 4.5 is fast for quick ideas. A free account covers everything here. If you write often, Claude Projects on the paid plans saves your instructions and voice sample once.
Start with this. Paste it at the top of a chat, or drop it in a Project.
Use it once per chat, as the base every later prompt builds on. Adapt by pasting your best posts, not your latest.
Prompts for hooks and openers
The first line decides whether the rest gets read, a big part of how the LinkedIn algorithm works.
Use it when you have a topic but no opener. Adapt by asking for hooks in the style of a past post that did well.
Use it when the opener buries the good part. Adapt by pasting the whole post so the rewrites match the payoff.