Step 2: Connect Google Sheets

Filter new rows, map the fields needed for content generation, and prepare the update back into the sheet.

1. In Filter Empty Records, keep the condition that checks Left Value as ={{ $json.LinkedIn_Post }} with the empty operator so only rows without a draft are processed.
2. In Map Row Fields, keep the JS code that maps Campaign, Subject, Audience, and rowNumber from the trigger data.
3. Configure Update Sheet Post with Operation set to update and Matching Columns set to ID.
4. In Update Sheet Post mapping, set ID to ={{ $('Sheet Row Monitor').item.json.ID }} and LinkedIn_Post to ={{ $json.output }}.
5. Credential Required: Update Sheet Post needs Google Sheets credentials, but none are configured—add your Google Sheets OAuth2 credentials before running.

Step 3: Set Up Trend Enrichment and Prompt Preparation

Enrich the post topic with trending LinkedIn results and format the AI prompt input.

1. In Trend Lookup, set Query to ={{$json["Campaign"] + " " + $json["Subject"] + " for " + $json["Audience"] + " site:linkedin.com"}}.
2. Credential Required: Connect your tavilyApi credentials in Trend Lookup.
3. In Compose Prompt Input, keep the JS code that creates chatInput using the Sheet fields and Tavily results.
4. Verify the execution flow: Map Row Fields → Trend Lookup → Compose Prompt Input.

Step 4: Set Up AI Drafting and Approval Logic

Generate the LinkedIn draft using the agent, route for approval, and handle revisions.

1. In Generate LinkedIn Draft, keep the Text prompt as defined, including the expressions like {{ $('Sheet Row Monitor').item.json.Campaign }} and {{ $json.chatInput }}.
2. Ensure Ollama Chat Engine is connected as the language model for Generate LinkedIn Draft. Add credentials to Ollama Chat Engine if your Ollama setup requires them.
3. In Send Approval Request, keep Operation as sendAndWait, Subject as Approval for LinkedIn Post Content, and the message body containing {{ $json['LinkedIn_Post '] }}.
4. Credential Required: Connect your gmailOAuth2 credentials in Send Approval Request.
5. In Approval Decision, keep the condition Left Value as ={{ $json.data['Do you approve this text?'][0] }} and Right Value as ={{"Yes"}}.
6. Confirm the branching behavior: Send Approval Request → Approval Decision, and if “Yes” it goes to Draft Image Prompt; if “No” it loops back to Generate LinkedIn Draft for revisions.

Step 5: Configure Publishing Outputs

Create the image prompt, generate the visual asset, and publish the LinkedIn update.

1. In Draft Image Prompt, set Model to llama3.2:latest and keep the message content ={{ $('Update Sheet Post').item.json['LinkedIn_Post '] }}.
2. Note that Ollama Tool Message is an AI tool sub-node connected to Draft Image Prompt; if tool credentials are needed, add them to Draft Image Prompt, not the sub-node.
3. In Create Visual Asset, set Resource to image and Prompt to =Generate an image for a linkedin post this is the description:{{ $json.content }} .The images should be realistic for linkedin..
4. Credential Required: Connect your openAiApi credentials in Create Visual Asset.
5. In Publish LinkedIn Update, set Authentication to communityManagement and Share Media Category to IMAGE.
6. Credential Required: Connect your linkedInCommunityManagementOAuth2Api credentials in Publish LinkedIn Update.

Step 6: Test and Activate Your Workflow

Run a full test to validate the data flow from the sheet to LinkedIn publishing, then activate for production.

1. Manually add a new row in the target Google Sheet with Campaign, Subject, and Audience populated, leaving LinkedIn_Post empty.
2. Use Execute Workflow in n8n to test the run and confirm Generate LinkedIn Draft creates a post and Update Sheet Post writes it back to the sheet.
3. Approve the draft via the email from Send Approval Request and verify Draft Image Prompt → Create Visual Asset → Publish LinkedIn Update completes successfully.
4. When the test is successful, switch the workflow to Active to run continuously.