How to Use n8n and AI to Automate Your Weekly Content Repurposing (Set It Up Once, Run It Forever)

You published a great blog post. Now you have to turn it into three tweets, a LinkedIn post, an email newsletter, and an Instagram caption — and you have to do it all over again next week.

That manual repurposing loop is one of the biggest time traps for solopreneurs. But here's the thing: it's completely automatable. With n8n and an AI model like Claude or ChatGPT, you can build a workflow that watches for new content and automatically generates every format you need — without you touching a keyboard.

This guide walks you through exactly how to build it.

Why n8n Is the Right Tool for This

You've probably heard of Zapier and Make. n8n is the third major workflow automation platform — and for AI-heavy use cases, it's often the best choice.

Here's why solopreneurs are switching to it:

Self-hostable: run it on a $5/month VPS and pay nothing per workflow execution
Native AI nodes: built-in connectors for OpenAI, Claude, and Ollama with no API wrangling
Code when you need it: drop into JavaScript or Python inside any node when logic gets complex
No execution limits: Zapier's free tier caps you at 100 tasks/month; n8n on your own server has no cap

For content repurposing specifically, n8n's ability to chain AI calls — generate tweet thread, then generate email, then post to Notion — without hitting per-step limits makes it dramatically more practical.

What This Workflow Does

Here's the end-to-end flow we're going to build:

Trigger: You add a new blog post URL to a Google Sheet (or Notion database)
Fetch: n8n scrapes the post content
AI Generate: Claude/ChatGPT produces five content formats simultaneously
Route: Each format goes to its destination (Notion draft, email draft, tweet queue)
Notify: You get a Slack or email message that everything is ready for review

Total time after setup: you paste one URL, and within 90 seconds you have a full week of content drafts waiting for your approval.

Step 1: Set Up Your Trigger

The cleanest trigger for most solopreneurs is a Google Sheets row. Every time you add a URL to a specific column, the workflow fires.

In n8n:

Add a Google Sheets Trigger node
Set it to watch for new rows in your "Content Queue" sheet
Map the URL column as your input variable

Alternative: If you use Notion for content planning, use the Notion Trigger node instead and watch for new pages in your "Blog Posts" database. Same logic, different source.

Set the polling interval to every 15 minutes — fast enough to feel instant, not so fast that it hammers the API.

Step 2: Scrape the Blog Post Content

Once you have the URL, you need the actual text. n8n has an HTTP Request node that handles this cleanly.

Add an HTTP Request node set to GET, pass in your URL variable, and set the response format to "Text". This returns the raw HTML.

Then add a Code node (JavaScript) to strip the HTML tags:

const html = $input.first().json.data;
const text = html
  .replace(/<style[^>]*>[\s\S]*?<\/style>/gi, '')
  .replace(/<script[^>]*>[\s\S]*?<\/script>/gi, '')
  .replace(/<[^>]+>/g, ' ')
  .replace(/\s{2,}/g, ' ')
  .trim()
  .slice(0, 8000); // keep first 8k chars for AI context

return [{ json: { cleanText: text } }];

This gives you clean, readable body text to feed into the AI. Capping at 8,000 characters keeps you within comfortable token limits for most models.

Step 3: Generate All Five Content Formats

Now the real work happens. Add an OpenAI or Claude node and write a single prompt that asks for all five formats at once. Getting everything in one API call is faster and cheaper than chaining five separate calls.

Here's the prompt template that works well:

You are a content strategist for a solopreneur business. 
Based on the blog post below, generate the following five content pieces.
Return ONLY valid JSON with these exact keys: tweet_thread, linkedin_post, email_subject, email_body, instagram_caption.

BLOG POST:
{{cleanText}}

REQUIREMENTS:
- tweet_thread: Array of 4-5 tweets, each under 280 chars, conversational tone
- linkedin_post: 150-200 words, insight-first, ends with a question
- email_subject: 3 subject line options, curiosity-driven, under 50 chars each
- email_body: 200-word email, personal tone, one clear CTA
- instagram_caption: 100 words, hook in first line, 5 relevant hashtags at end

Set the model to gpt-4o or claude-3-5-sonnet depending on your preference. Claude tends to produce more natural-sounding social copy; GPT-4o handles structured JSON output more reliably.

In n8n, enable "JSON Response" parsing so the output automatically becomes structured data you can route.

Step 4: Route Each Format to Its Destination

This is where n8n's power really shows. After the AI node, add a Split Out node to break the JSON into separate items, then route each one.

Three use cases to wire up:

Use Case 1: Tweets → Buffer or Typefully Add an HTTP Request node that POSTs to the Buffer API or Typefully API. Pass the tweet_thread array as the body. Your tweets are now queued for scheduling — you just need to approve and set a time.

Use Case 2: Email → Gmail Draft Use the Gmail node set to "Create Draft". Pass the email_subject (first option) and email_body. The draft appears in your Gmail ready to personalize and send — no copy-pasting from a doc.

Use Case 3: All Formats → Notion Create a Notion page in your "Content Drafts" database with each format in its own property block. This gives you one place to review and edit everything before anything goes live.

Step 5: Notify Yourself

Add a final Send Email or Slack node that fires when the workflow completes. The message should include:

The original blog post title
A direct link to the Notion draft page
A count of formats generated
Any errors (use n8n's error output branch)

This notification is your prompt to spend 10 minutes reviewing and approving — versus the 4 hours you'd have spent writing from scratch.

Real-World Results

Here's what this workflow looks like in practice for a typical week:

| Task | Without Automation | With This Workflow | |------|-------------------|-------------------| | Tweet thread (4 tweets) | 30 min | 0 min (AI draft in 90 sec) | | LinkedIn post | 45 min | 5 min review | | Email newsletter | 60 min | 10 min review | | Instagram caption | 20 min | 2 min review | | Total | 155 min | 17 min |

That's roughly 2.5 hours saved per blog post. If you publish twice a week, you're recovering 5+ hours every week — 260+ hours per year — from a workflow you set up once.

Common Issues and Fixes

The AI returns broken JSON: Switch from gpt-4o to gpt-4o-mini with a strict JSON system prompt, or add a Code node after the AI that uses JSON.parse() with error handling to catch malformed output before it breaks downstream nodes.

Content scraping fails on some URLs: Some sites block bots. Add a fallback branch that emails you to paste the content manually if the scrape returns less than 500 characters.

Duplicate triggers: If your Google Sheets trigger fires twice for the same row, add a "Set" node at the start that writes a timestamp back to the sheet, then check for that timestamp before proceeding.

Getting Started Today

You don't need to build the full five-format workflow on day one. Start smaller:

Build just the scrape → AI → Notion path first
Test it on three past blog posts
Add the email draft route next week
Add tweet scheduling once the other two feel stable

n8n has a free cloud tier and a generous self-hosted option. Most solopreneurs can run this entire workflow on the free plan until they're doing high volume.

If you want a library of workflow templates like this one — plus prompts, checklists, and tools that actually save time — grab the free AI Morning Brief at aiproductivitydaily.com/free-tools. It's a weekly digest of the most practical AI productivity moves for solopreneurs, delivered to your inbox every Monday morning.

Stop rebuilding the wheel every time you publish. Build the workflow once — and let it run.