n8n SEO automation.
Wire The SEO Agent into n8n with two nodes: a typed action node for keyword research, drafting, audits, and publishing, and a signed webhook trigger that fires when a page goes live. Build a content workflow that runs itself.
FREE TRIAL · CANCEL IN ONE CLICK · NODE SHIPPING TO THE N8N COMMUNITY- Base URL
- https://www.theseoagent.ai/api/v1
- API key
- sk_live_••••••••••••
- Webhook secret
- whsec_••••••••
What is the n8n SEO connector?
It is an n8n community node, n8n-nodes-theseoagent, that wires The SEO Agent into your n8n workflows. A declarative action node runs five operation groups over the public API (project, article, keyword-research, audit, job) and can be called by an n8n AI Agent as a tool. A separate webhook trigger node starts a workflow on article.published or audit.completed, with signed payloads. Drafting still runs the fact-check pass and the quality gate, so the article your workflow receives is cited and scored. The package is built and being submitted to the n8n community and npm. Pricing is a flat $99 a month. See the API reference for the underlying calls.
HTTP Request nodes and copy-pasted prompts.
Most people who automate SEO in n8n stitch it together from raw HTTP Request nodes and a generic LLM node. It works until it does not, and nobody owns the output quality.
- Hand-built auth headers that break on every key rotation
- A bare LLM node that hallucinates stats with no fact-check
- No quality gate, so weak drafts publish straight to your site
- No signed trigger, so you poll an endpoint on a timer and hope
- Brittle JSON parsing between every node in the chain
One typed node. The whole pipeline behind it.
The community node exposes the agent as a first-class n8n action: pick a project, kick off keyword research, generate an article, run an audit, or check a job. Credentials are configured once and reused.
- Typed operations with field hints, no raw header wiring
- Drafting runs the fact-check pass and the quality gate server-side
- A signed webhook trigger fires on article.published and audit.completed
- Usable as a tool inside an n8n AI Agent, not just a manual step
- Same public API documented at /docs/api, just wrapped as a node
Four reasons the node beats a hand-wired n8n chain.
Five operations over the public API, no header wiring.
The declarative action node covers the operations that matter in a content workflow: list and read projects, generate and fetch articles, run keyword research, fire an audit, and poll a job. Each one is a single call against the same public REST API you could hit directly, just exposed as typed fields with hints instead of hand-built JSON. The heavy lifting (live keyword data, fact-checked drafting, the quality gate) is the same SEO automation pipeline that runs inside the app.
Start a workflow the moment a page goes live.
The trigger node is a real webhook receiver, not a polling loop. It fires on article.published and audit.completed, and every payload is signed so you can verify it came from us before any downstream node runs. Use it to ping a channel when a draft is ready for review, sync the new URL into a sheet, or kick off an internal-linking pass. The signing scheme is the same one described on the REST API and webhooks page.
Hand the node to an n8n AI Agent as a callable tool.
Because the action node is declarative, an n8n AI Agent can call it as a tool. Give the agent a goal like “find three low-difficulty keywords for this site and draft one article,” and it picks the operation, fills the fields, and reads the result back. That turns The SEO Agent into a step inside your own automation rather than a separate dashboard you babysit. If you prefer a model-context-protocol surface instead, the MCP server exposes the same operations to any MCP-aware client.
The draft that lands has already passed the gate.
A bare LLM node in n8n will happily emit a confident, unsourced draft and publish it. This connector will not. Drafting runs the fact-check pass and the quality gate server-side, so the article your workflow receives is already cited and already scored above threshold. That is the same standard behind our native WordPress publishing, and it is included in the flat $99 a month plan, no per-call meter.
Four steps from credential to running workflow.
In your n8n instance, add the n8n-nodes-theseoagent community node from Settings, Community Nodes. The package is built and being submitted to the n8n community and npm. Until it lands there, the node config on this page is the canonical reference you can mirror with an HTTP Request node against the public API.
Create an API key in your account settings, paste it into the The SEO Agent credential in n8n along with the webhook signing secret. One credential powers both the action node and the trigger node.
Drop the action node onto the canvas and choose project, article, keyword-research, audit, or job. Each operation maps to one public API call. Chain a Schedule Trigger before it to generate on a cadence, or call it inline from an AI Agent node.
Add the trigger node to start a workflow the moment an article publishes or an audit completes. The payload is signed, so you verify it came from us before you act on it. Route it to Slack, a CMS step, or a spreadsheet.
The connector against raw HTTP plus an LLM node.
| FEATURE | THE AGENT | EVERYONE ELSE |
|---|---|---|
| Auth in n8n | One typed credential, reused | Hand-built HTTP headers |
| Fact-check pass before publish | ||
| Quality gate (refuses weak drafts) | ||
| Reacting to a publish | Signed webhook trigger node | Poll an endpoint on a timer |
| Callable by an n8n AI Agent | ||
| Cost model | $99/mo flat, included | Metered per LLM token |
Q.01Is the n8n node available to install right now?+
The package, n8n-nodes-theseoagent, is built and we are submitting it to the n8n community node registry and npm. Until it appears there, treat the configuration on this page as the canonical reference. You can reproduce the same calls today with an n8n HTTP Request node pointed at our public API, then swap to the node once it is verified.
Q.02What can the action node actually do?+
It exposes five operation groups over the public API: project (list and read your projects), article (generate a draft and fetch results), keyword-research (run research for a project), audit (kick off a site audit), and job (check the status of a long-running task). Each operation is one call with typed fields and hints.
Q.03How does the webhook trigger node work?+
It registers a webhook and starts your workflow when an event fires. Today it supports article.published and audit.completed. Every payload is signed with a shared secret you set on the credential, so you can verify the request originated from us before any downstream node runs.
Q.04Can an n8n AI Agent call this as a tool?+
Yes. The action node is declarative, which is exactly the shape n8n AI Agents consume as tools. An agent can choose the operation, fill the fields, and read the result, so you can describe an outcome in plain language and let the agent run the SEO step itself.
Q.05Do I still get the fact-check and the quality gate through n8n?+
Yes. Drafting runs the full pipeline server-side, including the fact-check pass that cites or rewrites claims and the quality gate that refuses to publish below threshold. The node returns the finished, gated article. n8n is the orchestration layer, not a shortcut around quality.
Q.06What workflows do people build with it?+
Two common shapes. Generate on a schedule: a Schedule Trigger every weekday kicks the action node to draft the next article, then a publish step pushes it live. React to a publish: the trigger node fires on article.published and routes the new URL to Slack, a sheet, or an internal-linking step. You can combine both in one workflow.
Q.07How is this different from just using an HTTP Request node and an LLM node?+
A raw HTTP plus LLM chain gives you no auth helper, no fact-check, no quality gate, and no signed events. You own all of that wiring and all of the output risk. The connector wraps authentication into one credential and runs the cited, gated pipeline behind a single typed node, so a weak draft never reaches your site.
Q.08What does it cost to run through n8n?+
The same flat $99 per month as the rest of the product. There is no per-call or per-token surcharge for using the n8n connector, the REST API, or the MCP server. The first three days are a free trial so you can wire up a workflow and watch an article ship before you commit.
Put a real content agent in your workflow.
Three days for a dollar. Create an API key, drop the node on the canvas, and watch a cited, gated article ship to your site. Cancel in one click if it is not pulling its weight.
FREE TRIAL · CANCEL IN ONE CLICK