Policy Engine

This module configures runtime policy enforcement for AI agents in Lattice Runtime.

Features

• Define authorization policies for agent actions
• Rate limiting and resource quotas
• Action allowlists and blocklists
• Audit logging for policy decisions

Usage

module "policy-engine" {
  source   = "registry.latticeruntime.com/lattice/policy-engine/lattice"
  version  = "1.0.0"
  sidecar_id = lattice_agent.main.id

  # Define policies
  policies = [
    {
      name   = "api-rate-limit"
      effect = "allow"
      actions = ["api:*"]
      conditions = {
        rate_limit = "100/minute"
      }
    },
    {
      name   = "deny-destructive"
      effect = "deny"
      actions = ["system:delete", "system:shutdown"]
    }
  ]
}

Inputs

Outputs