Automate Content Repurposing with AI: Turn One Idea into a Week of Posts

How do you repurpose content with AI automatically? You set up a lightweight n8n workflow that pulls a single piece of source material, asks an LLM to rewrite it for three channels (Twitter, LinkedIn, and a newsletter), and hands each version off to a scheduler. The result is a full week of brand-consistent posts that publish on autopilot.

What is AI content repurposing? AI content repurposing is the process of using generative language models to transform a single piece of text (or audio/video) into multiple, channel-specific formats while preserving the original brand voice.

What you need

Estimated build time: 2-3 hours (including API key creation and testing).

Step-by-step build for content repurposing with ai

1. Create API credentials - OpenAI: generate a secret key in the OpenAI Platform. - Twitter: apply for Essential access at the Twitter Developer Portal and copy the Bearer Token. - LinkedIn: register an app at the LinkedIn Developer site and note the Client ID, Client Secret, and OAuth 2.0 access token. - ConvertKit: locate the API key under Account Settings → Advanced. - Buffer: obtain a personal access token from Settings → Your Apps.

1. Spin up n8n

 docker run -d --name n8n -p 5678:5678 \
 -e N8N_BASIC_AUTH_ACTIVE=true \
 -e N8N_BASIC_AUTH_USER=admin \
 -e N8N_BASIC_AUTH_PASSWORD=strongpassword \
 n8nio/n8n
 

What this does: launches a self-hosted n8n instance with basic auth on port 5678.

1. Create a new workflow called "Weekly Repurposer".

4. Add a "Webhook" node (HTTP POST) that receives the original content. - Path: /repurpose - Method: POST - Expected JSON: { "title": "string", "body": "string" }

1. Add an "OpenAI" node to generate channel-specific copy. Configure three separate executions (one per channel) using the same prompt template:

 {
 "model": "gpt-4o",
 "messages": [
 {
 "role": "system",
 "content": "You are a brand-voice specialist. Rewrite the supplied text for the target platform while keeping the tone consistent."
 },
 {
 "role": "user",
 "content": "Original: {{ $json[\"body\"] }}\nPlatform: {{ $parameter.platform }}\nLength: {{ $parameter.length }} characters"
 }
 ],
 "temperature": 0.7,
 "max_tokens": 500
 }
 

What this does: asks GPT-4o to produce a version of the source text tailored to the selected platform (Twitter, LinkedIn, or newsletter).

1. Add three "Set" nodes to tag the output with platform (twitter, linkedin, newsletter) and a schedule_time (e.g., {{ $now.add(1, "day").toISO() }} for the next day).

7. Add a "Buffer" node for Twitter and LinkedIn posts. - Choose "Create a post". - Map text to the OpenAI output. - Set scheduled_at to the schedule_time from the previous node.

Note: Buffer's free plan allows 10 scheduled posts per channel, which is enough for a single week.

8. Add a "ConvertKit" node for the newsletter. - Action: "Create a broadcast". - Subject: {{ $json.title }} - Content: {{ $json.body }} (the AI-generated newsletter version).

1. Connect the nodes: Webhook → OpenAI (split into three branches) → Set → Buffer/ConvertKit → End.

1. Activate the workflow and test with a curl command:

 curl -X POST https://your-n8n-instance.com/webhook/repurpose \
 -H "Content-Type: application/json" \
 -d '{"title":"AI Automation Blueprint","body":"Learn how to turn a single blog post into a week of social content using AI."}'
 

What this does: triggers the whole pipeline, creating scheduled posts across all three channels.

11. Monitor and iterate - Use n8n's execution log to spot failures. - Adjust the OpenAI prompt length or temperature to fine-tune brand voice. - If you need more than 10 scheduled posts per week, upgrade Buffer or add a second n8n node that queues posts in a Google Sheet and publishes via a cron trigger.

Result: Once the webhook receives a single article, the workflow automatically produces a tweet thread, a LinkedIn carousel caption, and a newsletter draft, each queued for the next seven days.

Where this breaks

Most breakages are quota-related; budgeting API usage and adding simple guards (delays, token refreshes) keeps the pipeline reliable.