How do you auto-publish LinkedIn posts with AI?

Quick Answer: You can build a LinkedIn post auto-publisher by connecting Make.com with Claude (Anthropic's AI) to pull post ideas from a queue, draft content automatically, and publish directly to LinkedIn on a schedule. The entire workflow runs without manual input once configured, taking a post idea from a spreadsheet to published in minutes.

Building a LinkedIn content engine sounds like a full-time job. Most founders and marketers either post inconsistently or burn hours every week writing, formatting, and manually hitting publish. There is a better way.

This tutorial walks you through building a LinkedIn post auto-publisher using Make.com and Claude. You will set up a workflow that reads post ideas from a queue, uses Claude to draft the post, and publishes it directly to LinkedIn on a schedule. No copywriter required. No manual publishing.

What You Need Before You Start

Before building the workflow, get these accounts and assets ready:

• Make.com account (free tier works to start)
• Anthropic API key (for Claude access via API)
• LinkedIn account with a personal profile or Company Page you want to post to
• Google Sheets (or Airtable) to hold your post idea queue
• Basic familiarity with Make.com scenarios (helpful but not required)

Step 1: Set Up Your Post Idea Queue in Google Sheets

Your idea queue is the fuel for the whole system. Claude needs a prompt or topic to work from, so this sheet is where you feed it.

Create a Google Sheet with these columns:

Add 5-10 ideas to start. Keep the Status column as "Pending" for rows you want the system to pick up.

Step 2: Create a New Scenario in Make.com

Open Make.com and click Create a new scenario.

Your scenario will follow this logic:

1. Trigger on a schedule
2. Read the next "Pending" row from Google Sheets
3. Send the idea to Claude via HTTP request
4. Receive the drafted post back
5. Publish to LinkedIn
6. Update the row status to "Published"

Add your first module by clicking the + icon and searching for Google Sheets.

Step 3: Connect Google Sheets and Pull the Next Pending Idea

Select the Search Rows action in the Google Sheets module.

Configure it as follows:

• Spreadsheet: Select your idea queue sheet
• Sheet: Your ideas tab
• Filter: Set column "Status" equals "Pending"
• Limit: 1 (you only want to process one post per run)

This pulls the oldest pending idea each time the scenario runs. Make.com passes that row's data to the next module.

Step 4: Draft the Post with Claude via HTTP Module

Make.com does not have a native Claude module yet, so you call the Anthropic API directly using the HTTP module.

Add an HTTP: Make a Request module and configure it:

URL:

https://api.anthropic.com/v1/messages

Method: POST

Headers:

x-api-key: YOUR_ANTHROPIC_API_KEY
anthropic-version: 2023-06-01
content-type: application/json

Body (raw JSON):

{
  "model": "claude-opus-4-5",
  "max_tokens": 1024,
  "messages": [
    {
      "role": "user",
      "content": "Write a LinkedIn post about the following topic: {{Post Idea from Google Sheets}}. Tone: {{Tone from Google Sheets}}. Keep it under 250 words. Use short paragraphs. No hashtag spam. End with a question to prompt comments."
    }
  ]
}

Map the {{Post Idea}} and {{Tone}} fields from your Google Sheets module output. Make.com's variable picker handles this with a click.

Test this step before moving on. You should see a full drafted post returned in the HTTP response body under content[0].text. If you want to get better output quality from Claude for marketing tasks, it also helps to understand how prompt structure affects responses, as covered in this guide to Claude Skills for marketing.

Step 5: Extract the Post Text from Claude's Response

Claude returns a JSON object. You need to pull out just the post text before passing it to LinkedIn.

Add a Set Variable or Tools: Parse JSON module and map:

content[0].text

This gives you a clean string with the drafted post, ready to publish.

Step 6: Publish to LinkedIn

Add the LinkedIn module and select Create a Post.

Connect your LinkedIn account when prompted. Make.com handles the OAuth flow.

Configure the post:

• Author: Your personal profile or Company Page URN
• Text: Map the extracted post text from Step 5
• Visibility: PUBLIC (or CONNECTIONS if you prefer)

That is it. When this module runs, the post goes live on LinkedIn immediately.

Step 7: Update the Row Status to "Published"

You do not want the same idea published twice. Add a Google Sheets: Update a Row module at the end.

Configure it to:

• Find the row by its row number (mapped from Step 3)
• Set Status to "Published"
• Set Published Date to {{now}} (Make.com's current timestamp variable)

This closes the loop. The next time the scenario runs, that row is skipped.

Step 8: Set the Schedule Trigger

Go back to the start of your scenario and set the trigger module.

Click the clock icon and configure your schedule:

• Daily at 8:00 AM works well for consistent posting
• Or set it to run every weekday if you want Monday-Friday publishing
• You can also trigger it manually any time by clicking Run once

Make.com will fire the scenario at your chosen interval, pick up the next pending idea, draft it with Claude, and publish without you touching anything.

How to Improve the Workflow Over Time

Once the basic version is running, these additions make it significantly more useful:

Add a review step: Instead of publishing directly, send the Claude draft to a Slack channel or email for a quick approval before it goes live. Add a Slack module before the LinkedIn module and use a webhook to trigger publishing only after you click approve.

Include post format variety: Add a "Format" column to your sheet (e.g. "list", "story", "hot take") and include it in the Claude prompt. This stops every post sounding the same.

Pull ideas automatically: Connect the Google Sheets queue to a Notion database, a Typeform, or even a separate AI workflow that generates post ideas weekly. Your queue fills itself.

Track performance: After publishing, use Make.com to log the LinkedIn post URL back to your sheet. You can then manually add engagement data later for pattern analysis. If you are trying to tie LinkedIn activity back to pipeline or broader growth metrics, these B2B SaaS benchmarks can help give you context for what "good" looks like.

Why Make.com and Claude Work Well Together for This

Make.com handles the orchestration: reading data, calling APIs, writing back to sheets, and triggering LinkedIn. Claude handles the writing. Neither tool is trying to do both jobs, which keeps the workflow stable and easy to debug.

Claude produces LinkedIn-native copy well when given a specific topic and tone instruction. The output is not generic. Given a sharp prompt, it writes with a point of view, which is what LinkedIn's algorithm rewards.

The combination gives you a LinkedIn post auto-publisher that costs roughly $0.01-0.05 per post in API credits, takes under two hours to build, and runs indefinitely without maintenance.

FAQs

What is a LinkedIn post auto-publisher? A LinkedIn post auto-publisher is a workflow or tool that drafts and publishes LinkedIn content automatically, without manual input at the time of posting. Systems like the one in this tutorial use a content queue, an AI writing model, and an automation platform to move from idea to published post on a set schedule.

Is auto-publishing to LinkedIn against LinkedIn's terms of service? Publishing via LinkedIn's official API is permitted under LinkedIn's terms of service. Tools like Make.com use the official API with OAuth authentication. What LinkedIn restricts is scraping, fake engagement, and automated connection requests. Scheduled or automated post publishing through the API is explicitly supported and used by tools like Buffer and Hootsuite.

How is this different from using Buffer or a scheduling tool? Buffer and similar tools schedule posts you have already written. This workflow writes the post for you using Claude, then publishes it. The difference is the drafting step: you only need to supply a short idea or topic, not a finished post. That removes the biggest time cost in LinkedIn content creation.

Can I use this workflow for a LinkedIn Company Page instead of a personal profile? Yes. When configuring the LinkedIn module in Make.com, select your Company Page URN as the author instead of your personal profile. You need to be an admin of the page for the OAuth connection to work. If you are publishing to a page, it is also worth making sure the destination page converts traffic well, and these SaaS landing page best practices and examples are a useful reference.

What happens if Claude returns a post that is too long for LinkedIn? LinkedIn's character limit for posts is 3,000 characters. If you set max_tokens to 1024 in the Claude API call and instruct it to stay under 250 words, you will not hit the limit in normal use. As an extra safeguard, add a Make.com text length filter after the parse step that truncates or flags any response over 2,800 characters before it reaches the LinkedIn module.