n8n Inbox Triage with LLM Classification

Classify incoming email into categories with structured output, then label, route, or draft replies.

Nodes

1. Gmail Trigger (poll every minute) or Email Trigger (IMAP).
2. Basic LLM Chain with a Structured Output Parser using the schema below.
3. Switch on category.
4. Branches: add Gmail label, forward to a teammate, create a task, or Gmail → Create Draft for needs_reply.

Output schema

{
  "type": "object",
  "properties": {
    "category": {"type": "string", "enum": ["needs_reply", "fyi", "invoice", "newsletter", "spam_like"]},
    "urgency": {"type": "string", "enum": ["low", "normal", "high"]},
    "summary": {"type": "string"}
  },
  "required": ["category", "urgency", "summary"]
}

Prompt

Classify this email. Use only the defined categories.
From: {{ $json.from }}
Subject: {{ $json.subject }}
Body (truncated): {{ $json.text.slice(0, 4000) }}

Create drafts rather than sending automatically; review them before they go out.

Failure modes

• Misclassified urgent emails get buried; review low-confidence labels.
• Long threads exceed context; trim quoted history before classifying.
• Email content can carry prompt injection; never let triage take actions on its own.

Inspiration

Real-world reports that inspired or relate to this blueprint, credited to Agent Use Cases:

• Triage and draft replies for three email accounts (agentusecases.ai)
• Draft email replies and LinkedIn posts with n8n workflows (agentusecases.ai)

Markdown version: /blueprints/n8n-email-triage.md

Related blueprints