kubeswarm Security - Defense-in-Depth for Agents on Kubernetes

Defense-in-depth security for kubeswarm agent workloads on Kubernetes. Multiple overlapping controls ensure safe agent operation in production.

• Overview - Security layers and philosophy (this page)
• Guardrails and Trust - Tool allow/deny, trust levels, execution limits
• MCP Policy - URL allowlist, auth requirements
• Network Policies - Pod egress control
• API Key Management - Native Kubernetes Secrets