Netspective Logo
Netspective Content Engineering GuidelinesAI-Native Content Automation Pipeline

Node-Level Reference

Technical reference for whoever builds, reviews or maintains the workflow — node types, WordPress authentication, operational notes, and screenshot capture guidance.

For whoever builds, reviews or maintains this workflow.

Node-level architecture of the content pipeline in n8n

FIGURE 9 — Node-level architecture.

For the same graph drawn as the n8n canvas, alongside the workflow JSON it corresponds to, see The Workflow Artifact.

Nodes

StageNodeType
TriggerSchedule Triggern8n-nodes-base.scheduleTrigger
CollectRSS Readn8n-nodes-base.rssFeedRead
CollectHTTP Requestn8n-nodes-base.httpRequest
CollectHTML Extractn8n-nodes-base.html
CollectMergen8n-nodes-base.merge
CollectRemove Duplicatesn8n-nodes-base.removeDuplicates
DraftAI Agent@n8n/n8n-nodes-langchain.agent
DraftChat model (sub-node, ai_languageModel)e.g. @n8n/n8n-nodes-langchain.lmChatAnthropic
DraftStructured output parser (sub-node, ai_outputParser)@n8n/n8n-nodes-langchain.outputParserStructured
PublishWordPressn8n-nodes-base.wordpress

Node types verified against the node library bundled with n8n-mcp. The chat model is one option among several — any model n8n supports can occupy that slot.

Authentication

WordPress offers two credential types in n8n: wordpressApi (username plus application password) and wordpressOAuth2Api.

Prescribed: wordpressApi with an application password, issued to a dedicated WordPress account created for this purpose, holding the minimum role that can create posts. Application passwords are revocable individually without disturbing any human user's login, and scoping to a service account means the audit trail shows plainly which posts came from the pipeline.

Requirements: the WordPress REST API must be reachable from wherever n8n runs, and application passwords must be enabled (they are by default on current WordPress, but some security plugins disable them).

Operational notes

  • Deduplication state persists between runs. Confirm the retention window — long enough to avoid re-drafting last week's story, short enough not to grow unbounded.
  • Scheduling. The workflow arrives inactive on deployment and must be activated deliberately.
  • Error branch notifies rather than failing silently. Nominate the recipient.
  • Rate limits apply on both the AI model and the WordPress REST API. At one run per day neither is a concern; at hourly cadence, both need checking.
  • Credential rotation. The application password and model API key should be on whatever rotation schedule your policy requires; both can be replaced without rebuilding the workflow.

Screenshots to capture after build

Three screenshots complete this reference once the pipeline exists. Do not substitute mock-ups or stock images — an invented screenshot of a system that does not exist is the fastest way to lose a reader's trust in the rest of the document.

Capture at 1600px wide minimum, on a light background.

What to captureNotes
The completed workflow open in the n8n editorZoom so every node and connection is visible in one frame. Collapse the left sidebar. Make sure no credential names or instance URLs are legible.
WordPress Posts → Drafts, with at least one pipeline-generated article waitingUse a realistic headline, not "Test post 1". Crop out the admin bar if it shows the live site domain.
One generated draft open in the WordPress editorShow the headline, the opening paragraphs, and the tags/categories panel — this is what the reviewer actually sees.

Before sharing externally

  • No API keys, tokens or application passwords visible anywhere in frame
  • No internal hostnames, IP addresses or instance URLs legible
  • No real customer or third-party data in sample content
  • Author account name is the service account, not a personal login
  • Browser profile avatars and bookmark bars cropped out

How is this guide?

Last updated on

On this page