Most marketing teams learn their CAC doubled on Tuesday by reading the Monday dashboard next week. The 90-day pipeline that catches it within hours: GA4 + Meta + Google Ads into a Sheet, nightly Gemini call with a 14-day baseline and a name-the-cause prompt, red/yellow severity to Slack. Includes the actual prompt, the n8n JSON, and a real Tuesday alert.
An n8n + Claude Haiku 4.5 agent that watches 9 subreddits, scores every post on a 3-axis rubric (intent, urgency, fit), and posts a top-5 daily digest to Slack — with the verbatim Claude prompt, the 9-sub shortlist, the cost math, and the week-2 finding that pruned it down to 4 subs for a +120% signal lift.
A production n8n workflow that pulls decaying posts from GSC (Google Search Console, 谷歌站长工具), drafts a refresh brief, has ChatGPT write the update, validates the new draft, and ships it to the CMS — all in one loop. The full node map, the three prompts, the four validation gates, and the three failure modes that have actually cost me a post.
Wire a Slack slash command to an n8n workflow that fans out to Perplexity for cited research, hands the result to Claude for a 3-paragraph rewrite, and posts it back into the same channel — full JSON, prompts, and the dispatch_failed trap included.
Every Monday at 8:14 a.m. a Slack message lands with 4-9 underperforming posts. The n8n + Claude agent that writes it, the prompts that make Claude useful (and the ones that make it noisy), and the three things I changed after six weeks of running it for real.
A production-grade 4-agent content pipeline built in n8n. Each agent has a single job, a structured output, and a retry path. Includes the full workflow JSON you can paste into n8n, the prompts I use for each agent, the state schema that flows between them, and the failure modes that have actually cost me time.