Worker

Repurpose One Blog Post into a Week of Channels

A repurposing worker that scrapes a blog post or YouTube video with Firecrawl, then uses a LangChain chain to spin it into a LinkedIn post, a Twitter thread, and an email-newsletter section in one run.

What This Builds

This recipe builds a content multiplier. You give it one source — a published blog post or a YouTube video — and it produces several channel-native derivatives in a single run: a LinkedIn post, a Twitter/X thread, and an email-newsletter section. Firecrawl pulls the source into clean text, and a LangChain chain fans that text out into each format with channel-specific prompts. The output lands in a Google Doc (or your CMS) for a quick human edit before posting.

The Stack

  • Firecrawl scrapes the blog URL (or transcript page) into clean markdown so the chain reasons over structured text.
  • LangChain orchestrates the repurposing: one input, multiple parallel output chains each with a format-specific prompt template.
  • Anthropic | Claude generates each derivative in the right tone and length for its channel.
  • Google Docs API collects the generated pieces in one editable doc for review.

Step-by-Step Outline

  1. Accept a source URL (blog post or YouTube video) as input.
  2. Scrape it with Firecrawl to get clean markdown / transcript text.
  3. Build a LangChain chain with separate prompt templates per channel: LinkedIn post, Twitter thread, newsletter blurb.
  4. Run the chains over the same source text so each derivative stays faithful to the original.
  5. Write all outputs into a single Google Doc section, labeled by channel.
  6. A human reviews and edits in the doc, then copies each piece to its destination (or hands them to a scheduling agent).

Why This Shape Works

The expensive part of content marketing is not writing the first piece — it is the manual grind of reshaping it for every channel. Repurposing from one clean source keeps the message consistent while LangChain’s per-channel prompts handle tone and length differences. Landing everything in a review doc keeps a human in the loop, which is essential for brand voice and accuracy before anything is published.

Source

LangChain content-repurpose agent walkthrough and the n8n “Repurpose blog & YouTube content to social media” workflow: https://n8n.io/workflows/11599-repurpose-blog-and-youtube-content-to-social-media-with-gpt-51-and-google-docs/