Ultimate Human-in-the-Loop AI Content Engine for WordPress - n8n Workflow

Content Agencies scaling their output using AI.
SEO managers requiring highly structured, researched blog posts.
Bloggers or marketing teams who use Airtable for content pipeline management.
n8n users seeking advanced examples of Langchain Agents and complex n8n node structures.

This sophisticated n8n workflow solves the challenge of generating high-quality, researched, and structured content at scale while maintaining human oversight. It integrates Airtable as the central command center, coordinating various stages: research, drafting, image generation, and publishing. By utilizing AI agents from OpenAI and Perplexity for deep content research (even crawling a sitemap for internal link opportunities), this n8n automation ensures content is both relevant and optimized. It demonstrates the power of custom n8n templates for serious, production-level marketing operations, reducing manual bottlenecks and ensuring consistency across the entire publishing pipeline.

This comprehensive n8n workflow operates across five distinct phases, primarily orchestrated by dedicated n8n trigger nodes linked to status updates in Airtable:

1. Topic Initiation: The workflow starts with the Create New Topic n8n trigger (Airtable). It uses the Initial Research AI agent (powered by OpenAI and a Perplexity Search Tool) to rapidly generate a topic overview and chapter structure, saving the outline back to Airtable.


2. Chapter Planning & Research: Triggered by a new status, the flow retrieves website sitemap data (Get Post Sitemap) to gather internal linking context. A Switch Deep Research n8n node routes the flow based on research intensity. For standard research, the Researcher agent gathers data, which is then structured by the Copywriter chain. For deep research, the content is broken into sub-chapters, researched using dedicated AI nodes (Subchapter Researcher, Subchapter Copywriter), and aggregated.


3. Content Compilation & Image Generation: Upon chapter approval, the Airtable Select Content Trigger activates. The n8n workflow creates a Google Drive folder and uses HTTP Requests (likely DALL-E via OpenAI API) to generate and upload featured images and individual chapter images. It then merges all researched chapter data and converts the compiled content from Markdown to clean HTML (Markdown To HTML). The final HTML is saved back to the Airtable 'Finalize Post' table.


4. Image Upload Integration: A separate n8n trigger (Airtable Select Chapters Image Trigger) handles the upload and association of chapter images, ensuring all media is correctly linked within the content structure.


5. WordPress Publication: The Airtable Finalize Post Trigger initiates the final step. After confirming the status, the Post on Wordpress n8n node publishes the article. Subsequent HTTP requests handle setting the featured image and excerpt, concluding the n8n workflow by backing up the published post details and updating the status in Airtable.

To deploy this advanced n8n workflow, follow these steps:

1. Import the n8n JSON: Copy the provided JSON data and import it into your n8n instance via the 'Workflows' menu.


2. Airtable Setup: Ensure you have the necessary Airtable base structure defined (Topics, Chapters, Content, Final Posts). Set up credentials for Airtable in n8n and configure the four primary Airtable Trigger nodes (Create New Topic, Select Chapters Trigger, etc.) to monitor the correct base and table status changes.


3. AI Credentials: Configure your OpenAI (Chat Model) and Perplexity (Tool) credentials. These are crucial for the various Langchain n8n node instances.


4. Integration Setup: Configure the Google Drive, WordPress, and generic HTTP Request nodes with their respective credentials. The HTTP Request nodes are used for WordPress API calls (like media upload) and potentially for DALL-E/Perplexity interactions.


5. Webhook URLs: The workflow heavily relies on many Wait n8n node functions (e.g., Wait To Update New Topic). You must ensure your n8n instance is publicly accessible and that the Wait nodes can receive their respective webhook signals for asynchronous operation. This n8n template requires careful configuration of credentials to function correctly.

This complex n8n workflow utilizes dozens of interconnected nodes. Key n8n node components include:

Airtable Trigger & Node (Multiple Instances): Acts as the backbone, initiating phases and storing all content data (drafts, research, final HTML). Essential for the human-in-the-loop design.
Langchain Agent (Initial Research, Researcher, Subchapter Researcher): These advanced n8n nodes use LLMs (OpenAI Chat Model) combined with tools (Perplexity) to conduct web searches and structured research for content generation.
Langchain Chain (Copywriter, Subchapter Copywriter): Responsible for taking raw research output and drafting cohesive, structured content based on the findings.
Structured Output Parser: Used alongside AI nodes to ensure the generated content (e.g., chapter outlines) is returned in a predictable JSON format for subsequent n8n node processing.
HTTP Request Node: Vital for complex operations, including retrieving sitemap XML data, generating images via external APIs (like DALL-E), and handling custom WordPress REST API calls (featured image upload).
WordPress Node: The final publishing n8n node, taking the generated HTML content and publishing it to the linked WordPress site.
Code Nodes (New Chapters, Parse Chapters, Combine Into Article): These run custom JavaScript to transform and structure data outputs from the AI agents into the desired format before writing to Airtable or compiling the final post.
If and Filter Nodes: Critical for flow control, ensuring the n8n workflow only proceeds when Airtable status fields meet specific 'Confirmed' or 'In Progress' criteria.