n8n automation kits — build 2026.01
Three production-grade n8n workflows for content repurposing, personalised outreach and RAG support — with the documentation you wish templates came with.
A screenshot of eleven connected nodes, a title promising an AI agent, and a JSON file that breaks the moment you import it.
You know the pattern. Half the nodes reference credentials that were never abstracted. The prompt is one line. There is no error handling, so the first unreachable URL kills the run. There is no documentation, so you end up reverse-engineering someone else's shortcuts.
You didn't want a template. You wanted the two hours back.
Point it at your YouTube channel, blog or podcast feed. Every morning it takes what you published and writes the LinkedIn post, the X thread, the short-form script and the newsletter — then logs all of it to a content calendar and leaves the newsletter as a Gmail draft. You approve, you publish.
Drop leads in a sheet. It reads each company's actual homepage, writes a first email that opens with something specific and true about that business, scores the fit from 1 to 5, and drafts only the ones worth your time. It never sends. Drafts only, on purpose.
Load your FAQs and docs into a sheet. It chunks, embeds and stores them, then answers questions strictly from that knowledge — and says it does not know instead of making something up. Short conversation memory included.
It runs on import. Placeholders are marked. Credentials are the only thing you add.
It fails safely. A dead link, an empty feed, a site behind bot protection — none of them crash a run. That handling is in the file, not in a “you should probably add” note.
It explains itself. Sticky notes inside every workflow tell you which nodes need attention and which two or three numbers you'll actually want to tune. Plus a setup guide covering the errors people hit, in the order they hit them.
Known limitations — stated before you pay
One price, one download, yours for good.
If a workflow doesn't import and run as described, and setup support doesn't fix it, you get your money back. No form, no window, just ask.
Yes, Cloud and self-hosted. Tested on n8n 2.31. Requires 2.x — the AI nodes are used at versions that 1.x releases will not import, so this is a hard requirement, not a suggestion.
You pay the model providers directly — we take nothing. Workflow 01 uses roughly 3-6k tokens per item processed, which we measured; three items a day is a few dollars a month. Workflows 02 and 03 scale with your own volume, so your bill is whatever you configure them to do. Bring your own Anthropic key, plus an OpenAI key for the RAG workflow's embeddings.
Yes — that's the main use case. Deploy them for as many paying clients as you like. The only restriction is that you can't resell or republish the files themselves.
Yes. Replace the model node — it is a sub-node, so it is a drop-in swap. The setup guide includes a tested walkthrough for running the whole pack on a single OpenRouter key instead of separate Anthropic and OpenAI accounts. The prompts are written against Claude's instruction-following and will need a light pass if you move to a weaker model.
Yes. Fixes and improvements to these three workflows are free for life — you bought the pack, not a snapshot of it. New workflows will be sold as separate packs.
Email support for setup, answered within two business days. It covers getting the product working as described — not building custom workflows for you.