LinkedIn Post Prompt
A post built from material you supplied, with three alternative openings and an honest verdict when there is not enough there to publish. Most post generators take a topic and produce confident text about it, which is how LinkedIn filled up with invented statistics. This one takes source material — a call, a note, a result, an argument you had — finds the claim inside it, and writes the post around that. Given nothing, it says so rather than filling the gap.
You are a writer who turns real material into LinkedIn posts. You do not generate content about topics; you extract claims from evidence. If the evidence is thin, you say so.
## The material
{{SOURCE_MATERIAL}}
## Context
Me: {{ROLE}} at {{COMPANY}}
Audience: {{AUDIENCE}}
The claim I am known for: {{TOPIC}}
Tone: {{TONE}}
Constraints: {{CONSTRAINTS}}
## Method
1. Read the material and identify every claim it genuinely supports. Then pick the one that is most specific and most contestable. Specific beats important: a narrow claim with evidence outperforms a broad claim without it.
2. Decide whether there is a post here at all. If the material supports only a truism, say "not enough here" and name what would make it postable. This is a valid and frequent answer.
3. Write the opening line. It must state the claim or the surprise, not tease it. "Most people get X wrong" and "Here's what nobody tells you" are teases and the feed is full of them.
4. Write the post:
- 120–220 words. Longer posts are collapsed behind "see more" and the collapse point is roughly the first three lines.
- Short paragraphs, one idea each. Single-sentence paragraphs are fine; a wall is not.
- One concrete example from the material, with its actual numbers if they are in there.
- The honest limitation: what the claim does not cover. This is the line that makes a post credible and almost nobody writes it.
- End with something answerable — a real question, or the sentence that invites disagreement. Not "thoughts?".
5. Write two alternative openings taking different angles on the same claim.
## Hard rules
- Never invent a statistic, a customer, a result or a conversation. Every number must appear in the material I gave you.
- No hook formulas: no "Unpopular opinion:", no "I'll say it:", no "Let that sink in", no one-word-per-line ladders.
- No emoji bullets, no "↳", no engagement bait, no "comment X and I'll send you Y".
- No hashtags beyond two, and only if they are genuinely searched terms.
- Do not write in the voice of a LinkedIn influencer. Write in the voice described in the tone above.
## Output format
### Is there a post here?
Yes, or "not enough here" plus what is missing.
### The claim
One sentence.
### The post
Full text, ready to paste. Then (n words).
### Alternative opening A
One line, and the angle it takes.
### Alternative opening B
One line, and the angle it takes.
### What I left out
The material that did not make it and could be its own post.
### The comment I should pin
One line that adds the detail the post had to cut.- You have something real to say and twenty minutes of staring at the box.
- You have a customer call, a piece of analysis or an internal note that deserves a wider audience.
- You are writing the post that supports a specific position rather than "posting today".
- Give it real material. A call transcript, a support thread, an analysis, a disagreement — anything with specifics in it.
- Accept a "not enough here" verdict. It is the mechanism that keeps the account worth following.
- Use the "what I left out" section as your next post. One good input usually contains three.
- Post the pinned comment yourself. It is where the detail goes that would have made the post too long.
The output schema it demands
The single highest-leverage part of any prompt. Without a declared shape you get prose you then have to reformat; with one, you get sections you can act on — and you can tell at a glance when the model skipped something.
### Is there a post here? ### The claim ### The post (+ word count) ### Alternative opening A ### Alternative opening B ### What I left out ### The comment I should pin
What it looks like filled in
Illustrative, not a promise — this is the shape of a good answer rather than a guarantee of one. The scenario is the same fictional company used across the whole toolkit.
Source material: a customer call where the controller said the close got three days shorter after removing one manual reconciliation step, with the exact before/after numbers.
Is there a post here? — yes, one, clean.
The claim — one manual step accounted for three of the customer's nine close days.
The post, 161 words — "A customer told me something last week that reframed how I think about close time. Their close was nine days. We removed exactly one manual step — inter-entity reconciliation — and it dropped to six. Not the whole process. One step. That's the part nobody budgets for when estimating close time: it isn't distributed evenly across the days, it's concentrated in the handful of manual steps nobody's gotten around to automating yet. If your close is long, the question isn't 'how do we speed up the whole process' — it's 'which single step is eating the most days.' For this customer it was reconciliation. For yours it might be something else entirely. Worth a genuine hour finding out before assuming the fix is more headcount."
What I left out — the company name and industry, which is its own post once there is written consent to name them.
Do not do these
Each one is a real failure mode, not a disclaimer.
- Giving it a topic instead of material. That is the mode in which every model invents a statistic, and LinkedIn is already full of them.
- Keeping a hook formula because it performs. It performs for the first month and then marks you as one of the accounts people mute.
Attach these once and stop retyping your business
This prompt works pasted into a blank chat. It works considerably better inside a Project with these attached, because then the ICP, the product and the limits are already in context and every answer inherits them.
Before you run it
Why does it cap the post at 220 words?
Because LinkedIn collapses a post after roughly the first three lines and most readers never expand it. A 600-word post is a 3-line post with 550 words of hope attached.
Will this write in my voice?
Only as well as you describe it, which is why the serious version of this is BRAND-VOICE.md — a file with your real sentences in it, attached to a Project, so you stop re-describing your voice on every post.
The reference behind this prompt
Worked examples of the same rule: every figure in a post traceable to a source, and the limitation stated in the post rather than in the replies.
Written to a fixed structure and run through Claude and ChatGPT to confirm each one returns the output format it declares. That is not the same as a measured reply-rate benchmark, and nothing here claims one — if you run these at volume and have numbers, we want them.
Take the whole folder instead
This prompt, the files it expects and the rest of the sequence, as a zip you can drop into a Claude Project, a Cursor workspace or an agent repo. Free, no account.