Slack to Kubernetes: Claude Manifests with Guardrails in n8n

Generate safe Kubernetes manifests from Slack requests—powered by Claude and enforced guardrails in n8n

This n8n workflow turns a plain-English Slack request into a Kubernetes resource manifest using Anthropic Claude, validates it with deterministic safety guardrails, and requires human approval before applying it. It then checks rollout health and performs an automatic rollback if things go wrong.

What this workflow does

• Triggers from Slack or manual run: Listens for a Slack app mention (or supports manual execution) and captures the resource description to create.
• Generates a Kubernetes manifest with Claude: Uses Anthropic Claude to produce a structured Kubernetes manifest (as JSON) plus a short explanation.
• Validates with deterministic guardrails: The workflow parses the output and checks for safety requirements such as:
  • Allowed namespaces
  • Required CPU/memory requests and limits
  • Blocked usage like privileged, hostNetwork, or hostPath
• Stops on violations: If validation fails, it posts the specific violations back to Slack and halts.
• Human approval gate: If validation passes, it posts the manifest and explanation to Slack and waits for an approve/reject decision.
• Applies and verifies rollout: On approval, sends the manifest to the Kubernetes API, checks status/readiness, and if rollout health is not good, deletes the created resource and alerts Slack about the rollback.

Use cases

• Standardize and approve Kubernetes changes requested by teams via Slack.
• Add policy enforcement (namespaces, resource requirements, and disallowed settings) before anything reaches the cluster.
• Reduce deployment risk by requiring manual approval and automatically rolling back unhealthy outcomes.

Technical details

• Integrations: Slack (bot/token for mentions and posting), Anthropic (Claude), Kubernetes API via HTTP.
• n8n nodes/logic: if, set, code, wait, slack, and sticky note for workflow clarity.