n8n Webhook-Triggered AI Agent

Expose an n8n AI Agent (with memory and tools) behind a webhook that returns the answer synchronously.

Nodes

1. Webhook (POST /agent, Respond: Using 'Respond to Webhook' node).
2. AI Agent (@n8n/n8n-nodes-langchain.agent), prompt from {{ $json.body.message }}.
   Sub-node Chat Model (OpenAI, Anthropic, Ollama, ...).
   Sub-node Window Buffer Memory keyed on {{ $json.body.session_id }}.
   Tool sub-nodes, e.g. HTTP Request Tool, Calculator.
3. Respond to Webhook – JSON { "reply": "{{ $json.output }}" }.

Test

curl -X POST https://YOUR-N8N/webhook/agent \
  -H 'Content-Type: application/json' \
  -d '{"session_id":"u123","message":"What is 17% of 2,340?"}'

Hardening

• Add Header Auth on the Webhook node (shared secret) before exposing it publicly.
• Set a max iterations limit in the AI Agent options.
• Use the test URL (/webhook-test/) during development, and activate the workflow for the production URL.

Failure modes

• Unauthenticated webhooks let anyone spend your tokens; add header auth.
• Agent loops on tool errors without an iteration limit.
• Test vs production webhook URLs differ; activate the workflow before going live.

Markdown version: /blueprints/n8n-webhook-ai-agent.md

Related blueprints