Agent → Run Agent
AI-generatedSummary
Runs one selected Markup AI agent on the supplied document content, waits for the run to finish, and returns the agent result with issue counts and an HTML report. The node requires Markup AI API credentials and fails if the run does not complete successfully or exceeds the polling timeout.
Inputs
- Agent (required) — Select one agent to run. Terminology and AI Voice Detector agents can use Domain IDs; the Style Agent can use a Style Guide.
- Content (required) — Document text to analyze.
- Domain IDs — Optional terminology domains passed to the Terminology or AI Voice Detector agent.
- Style Guide — Optional configured style guide passed to the Style Agent. Markup AI auto-detects the content profile.
- Additional Options → Document Name — Human-readable document name included in the generated report.
- Additional Options → Document Reference — Caller-supplied tracking identifier. Sent to the agent and returned as document_ref.
- Additional Options → Timeout (Ms) — Maximum time to poll for completion, in milliseconds; defaults to 120000. Polls every two seconds when the initial response is non-terminal.
Output shape
One n8n item per input item, each containing a JSON object on success.
Successful output includes workflow_id, status, document_ref, result, started_at, completed_at, duration_seconds, issue_counts (total, high, medium, low), and html_report. A non-completed terminal status or polling timeout raises an error; with Continue On Fail enabled, the item instead contains {"error": "..."}. The operation also checks that the organization has the Style Agent enabled before processing items.
Examples
Example 1: Run Terminology on CMS copy
Select the Terminology agent, add one or more Domain IDs, provide the document text, and optionally set Document Name and Document Reference.
Example 2: Check writing against a configured style guide
Select the Style Agent, choose a Style Guide, provide the document text, and optionally set a tracking reference.