Production AI Playbook: Deterministic Steps & AI Steps (5 of 5)
Needs changesNeeds changes. This template uses 1 thing(s) that n8n 3.0 removes: AI Agent v1 (old agent modes). 1 breaking item(s) found. Static check, last run 2026-10-19.
What to replace before upgrading
| Removed in 3.0 | Replace with |
|---|---|
| AI Agent v1 (old agent modes) AI - Classify + Draft (v1.7) | Latest AI Agent node (Tools Agent). For SQL Agent, use a Postgres or MySQL tool with a recent AI Agent. |
Nodes used (9)
- Respond to Webhook
n8n-nodes-base.respondToWebhook - Code
n8n-nodes-base.code - Switch
n8n-nodes-base.switch - If
n8n-nodes-base.if - AI Agent
@n8n/n8n-nodes-langchain.agent - Webhook
n8n-nodes-base.webhook - Guardrails
@n8n/n8n-nodes-langchain.guardrails - OpenAI Chat Model
@n8n/n8n-nodes-langchain.lmChatOpenAi - OpenRouter Chat Model
@n8n/n8n-nodes-langchain.lmChatOpenRouter
What it does
The full end-to-end workflow that chains all patterns together. This template processes customer feedback from intake to team routing, with normalization, validation, native guardrails, AI classification, and confidence-based branching at every step. What you'll do - Send customer feedback through a webhook and watch it flow through every stage. - See the data get normalized, validated, and scanned by n8n's native Guardrails node for jailbreak attempts and PII. - Watch the AI classify feedback (bug report, feature request, praise, complaint, question) with a confidence score and generate a personalized response draft. - See AI-generated responses pass through output guardrails that check for NSFW content and secret keys before reaching users. - Watch high-confidence results route automatically: bug reports and feature requests to the product team, complaints to customer success, and praise to marketing as testimonial candidates. What you'll learn - How to chain normalization, validation, native guardrails, AI, and routing into a single pipeline - How to use n8n's Guardrails node for both input screening (jailbreak, PII, secret keys) and output screening (NSFW, secret keys) - How confidence-based branching separates high-confidence results from items that need human review - How Switch nodes route classified feedback to the right destination (product team, customer success, or marketing) - How every step between AI nodes is deterministic and inspectable - How all these patterns work together in a production-ready workflow Why it matters This is the complete picture. Individual patterns are useful on their own, but the real power comes from combining them into a pipeline where AI handles the judgment calls and everything else follows explicit, testable rules. Import this template as your starting point and connect your own integrations. This template is a learning companion to the Production AI Playbook, a series that explores strategies, shares best practices, and provides practical examples for building reliable AI systems in n8n. https://go.n8n.io/PAP-D&A-Blog