Content Repurposer (n8n)

What it does & who it's for

Content Repurposer is a production-ready n8n workflow that takes one long-form post and does the repurposing pass you keep postponing. A new article arrives by hourly RSS polling or a webhook POST (a URL, or raw text straight from your CMS). It is fetched and boilerplate-stripped into clean article text, then three LLM prompts draft the platform variants: a LinkedIn post (150-220 words, hook-first, no links in body), a numbered X thread (exactly 6 tweets, each under 240 characters, link only in the closer) and a newsletter blurb (labelled subject, preview and 60-90 word blurb, parsed into structured fields). The drafts land in your inbox with Approve and Reject buttons; nothing publishes without your click. On approval — and after a configurable undo window — three branches fire in parallel: the LinkedIn node posts to your profile, the Buffer branch queues each tweet of the thread in order, and an HTTP branch delivers the full draft pack JSON to your newsletter tool or scheduler. It is built for founders, solo marketers and content teams who publish long-form and then lose an hour per post rewriting it for each channel — or worse, paste the same link everywhere. The prompts are the actual product: each one encodes editorial rules that make output publishable rather than obviously generated (word bands, hook limits, platform link etiquette, labelled output that machines can parse), plus an anti-hallucination guard so a variant never invents a statistic you didn't write. All three are embedded in the nodes and mirrored in prompts.md with the rationale behind every rule, so you can adapt the voice without breaking the machinery. Everything is the real n8n export format: 19 typed nodes (webhook and RSS triggers, code, IF, HTTP request, the built-in OpenAI node, SMTP email, restart-safe Wait nodes, LinkedIn) plus four sticky notes documenting each stage on the canvas. No community nodes. Credentials are named placeholders you connect once after import. The approval flow uses n8n's own resume-URL mechanism, so parked executions survive restarts and each article waits independently. Structural guardrails fail loudly instead of publishing garbage: empty LLM output stops the run before the approval email, and the thread splitter refuses any tweet over 280 characters. You could wire something like this yourself — what you are buying is the edge-case work already done: dual triggers normalised to one shape, full-text feeds and raw-text posts that skip the fetch, HTML cleaning that survives real blog markup, prompts with checkable output contracts, a reject path that provably publishes nothing, and a setup guide that goes from import to live in 30-45 minutes with a zero-dependency test payload included. Run the interactive demo first — it replays the exact node logic and sample outputs shipped in the download, including the approve and reject paths.

Version history

1. v1.0Latest29 July 2026

Licence & terms

These terms are snapshotted at purchase — what you see here is what applies to your order.

Reviews