LinkedIn post and image generator

What it is

This n8n workflow turns a topic and a target audience into a finished LinkedIn post and a matching graphic, and emails both to the address you type in. A form takes three fields: the email address, the topic and the audience. An agent named LinkedIn Post Agent searches the web with Tavily, then writes the post with GPT-4.1 through OpenRouter. A second agent, Image Prompt Agent, rewrites the post as a prompt for a marketing-style graphic. OpenAI’s image model draws it, and Gmail sends the post. A LinkedIn node sits on the canvas with no line into it, for anyone who would rather publish than email.

Who it suits

Founders, consultants and marketers who post on LinkedIn often and lose time to the blank page, and the social media managers and small agencies who write posts for other people.

You need some n8n: importing a file, adding four credentials, reading a node’s output. It writes from a brief, not from your past posts, so skip it if you want a draft in a voice learned from your own writing. Its author puts setup at about 21 minutes, which assumes the four accounts already exist.

Where to use it

Any n8n you run will do, as long as it can reach OpenRouter, Tavily, OpenAI and Gmail.

• n8n Cloud: import the file into a new workflow there
• Your own n8n server: the same file, the same steps
• LinkedIn: wire the unused LinkedIn node in after Convert to Binary to publish without the email
• Another channel: replace Send Post with the node for Slack, Telegram or a drive

Setup

Get it saves a zip with workflow.json and Mindket’s setup notes in your downloads folder. Have these ready before you import it.

1. In n8n, create a workflow, open its menu, choose Import from File and pick workflow.json.
2. Add an OpenRouter credential and select it in GPT_4.1. Both agents use that node.
3. Add an HTTP Header Auth credential for Tavily: header Authorization, value Bearer and your key. Select it in Tavily.
4. Open Generate Image and put your key in the header where it says YOUR_KEY. Do the same in the test node Generate Image1.
5. Add a Gmail OAuth2 credential and select it in Send Post.
6. In Send Post, name the attachment’s binary property data. The file leaves it empty, and data is where the image is.
7. Save, send a test entry through the form’s test address, then press Publish.

The test chain at the bottom of the canvas, When clicking ‘Test workflow’, draws one image from a fixed prompt. It checks your OpenAI key without a whole run. Delete it once the main flow works.

How to use it

Open the form from On form submission and fill in your email address, a topic and a target audience. The agents then search, write, plan the graphic and draw it, and Gmail sends the mail. If nothing arrives, open the run under Executions to see which node stopped.

Example

This entry in the form:

Email: you@example.com
Topic of Post: Four-day work weeks
Target Audience: HR directors

LinkedIn Post Agent searches Tavily first, then writes a post that opens on a hook, cites its sources, carries a few hashtags and ends by asking readers what they think. Image Prompt Agent reads that post and describes one marketing graphic, with the post’s figures in it where it has any. The mail holds the post text and the graphic.

For newcomers and experts

If you are new. n8n draws a workflow as boxes joined by lines, and each box, a node, does one job. An agent is a node that lets a language model choose when to use a tool, here a web search. Run the form from its test address first. Every node then shows what it received and what it sent, which is where a wrong key turns up.

For experts. One chat model serves both agents, and the image step is a plain HTTP request.

Good to know

Each run costs money, and the form is open to anyone who has its address.

• Billing: OpenRouter bills two GPT-4.1 calls, Tavily bills the search and OpenAI bills the image
• Open form: once the workflow is published, anyone with the form’s address can start a run on your keys. The result is mailed from your Gmail to any address they type, so turn on the form trigger’s basic authentication or keep the address to yourself
• The key in the node: Generate Image keeps its header in the node, so the key is saved in the workflow and goes with any export. Move it into a credential before you share it
• Your data: the topic and audience go to OpenRouter, its model provider and Tavily, the image prompt goes to OpenAI, and the post and graphic go to the address in the form
• Facts: the search asks for one result, so a claim can rest on a single page. Read the post and its sources before it goes out
• Credentials: the JSON holds none, and nothing from its author’s account