My first n8n agent that saved me 5 hours a week

I was spending roughly 5 hours a week repurposing content. One idea, written four times for four formats.

It's the kind of work that feels productive but isn't. You're not thinking — you're rewriting. So I automated it.

The problem

My content flow looked like this:

1. Write a LinkedIn post (30 min)
2. Rewrite as a Twitter/X thread (45 min)
3. Extract a paragraph for the newsletter (20 min)
4. Expand into a blog post draft (60 min)

Every week. Same formats. Same cognitive overhead.

The solution: one post, four outputs

Now I write the LinkedIn post. I paste it into a form. n8n does the rest.

Here's the full flow.

Step 1: Trigger

A simple webhook in n8n. I have a tiny form (Notion form, or even a plain curl command) that sends the post content to the webhook URL.

Alternatively: a Gmail trigger that watches for emails from myself with subject [CONTENT]. Whatever is lowest friction for you to actually use.

Step 2: Twitter thread (Claude)

You are a content strategist. Convert this LinkedIn post into a Twitter/X thread.

Rules:
- First tweet: hook that doesn't need the rest to make sense
- 6-10 tweets max
- Each tweet under 280 characters
- Last tweet: call to action or summary
- No hashtags
- Preserve the original voice

Post: {{$json.post}}

The output is a JSON array of tweet strings. n8n parses it.

Step 3: Newsletter paragraph (Claude)

Extract the single most valuable insight from this post and rewrite it as a paragraph (3-5 sentences) for a weekly newsletter.

- Write in first person
- Conversational tone, not formal
- End with one concrete takeaway the reader can apply today

Post: {{$json.post}}

Step 4: Blog draft (Claude)

Expand this LinkedIn post into a structured blog post draft.

Structure:
- Title (SEO-friendly, not clickbait)
- Hook paragraph (why this matters)
- 3-5 H2 sections that develop the ideas in the post
- Conclusion with actionable next step

Write in markdown. The draft should be ~600 words. Don't pad with filler — if an idea doesn't have enough to say, drop it.

Post: {{$json.post}}

Step 5: Save to Notion

All four outputs go into a Notion database with properties:

• Status: Draft
• Source post (the original LinkedIn text)
• Thread (the tweets)
• Newsletter (the paragraph)
• Blog draft (the markdown)

From there I review, edit, and publish manually. The drafts aren't final — they're 70% of the way there, which is exactly what I needed.

What the time looks like now

Before: ~5h/week repurposing. After: 30 min writing the post + 15 min editing outputs = 45 min total.

The 5 hours went to 45 minutes. The difference is editing vs. writing from scratch — my brain works faster when there's something to react to.

What to watch out for

Voice drift: Claude will sanitize your voice if you're not explicit. Include examples of your writing in the prompt system message if consistency matters.

Blog drafts need real editing: The structure is good, but the specificity isn't. You need to add your actual examples, numbers, and context. Don't publish the draft as-is.

Webhook security: Add a secret header to your webhook and validate it in n8n. You don't want random people triggering your automation.

The setup

• n8n (self-hosted on Railway, ~$5/month) or n8n Cloud
• Claude API (Anthropic) — a month of content repurposing costs me under $3
• Notion API for the database writes

Total cost: under $10/month. Time saved: 4h+/week. The ROI math isn't complicated.

---

For the architecture behind this: Building AI agents that actually work — patterns that hold up week after week.

Still fuzzy on whether this is a "workflow" or an "agent"? The real difference between a prompt, a workflow, and an agent.