Automated Blog Publishing Pipeline

End-to-end n8n workflow that takes a content brief, drafts a full blog post with an LLM, runs it through an editing pass, generates SEO metadata, and publishes directly to the CMS.

n8nClaude APICMS APISEOContent Automation
2024-06

Overview

Producing consistent, SEO-optimised blog content at scale is expensive. This pipeline automates everything after the editorial brief — from first draft to live post — while keeping a human review step in the loop before publishing.

Workflow Stages

Stage 1 — Brief Intake

A content manager fills a simple form (title, target keyword, outline bullet points, desired tone). The form triggers the n8n workflow via webhook.

Stage 2 — Draft Generation

Claude receives a structured system prompt defining the site's style guide and voice, then generates a full post — headings, body copy, internal link placeholders, and a CTA.

Stage 3 — Editing Pass

A second LLM call with a different system prompt acts as a copy editor: it checks for passive voice overuse, keyword stuffing, heading hierarchy issues, and readability score. Suggestions are returned as structured annotations.

Stage 4 — SEO Metadata

A dedicated function node generates the <title> tag (≤60 chars), meta description (≤155 chars), Open Graph tags, and a suggested URL slug — all optimised for the target keyword.

Stage 5 — Human Gate

The draft + annotations + metadata are compiled into a review document and sent to a Slack channel. An approval/reject button triggers the next step via a webhook back to n8n.

Stage 6 — CMS Publish

On approval, the workflow calls the CMS REST API to create and publish the post, upload featured image metadata, and set the scheduled publish time.

Outcome

Brought blog publishing from a 3–4 hour manual process per post down to ~20 minutes (mostly the human review stage).


← BACK TO AI AUTOMATION & AGENTS