{
  "id": "skill-area:AI-agent-guardrails",
  "_kind": "SkillArea",
  "_file": "domain/skill-areas/skill-areas-expanded.yaml",
  "_cluster": "domain",
  "attributes": {
    "displayName": "AI Agent Guardrails",
    "description": "Runtime safety controls for AI agents — output filtering,\ntool permission boundaries, budget and rate limits, human-in-the-loop\ngates, and containment policies for autonomous systems.\n",
    "expertiseLevels": [
      "intermediate",
      "expert"
    ]
  },
  "outgoingEdges": [
    {
      "from": "skill-area:AI-agent-guardrails",
      "to": "specialization:ai-agents-conversational",
      "kind": "applies_to",
      "attributes": {
        "confidence": "primary"
      }
    },
    {
      "from": "skill-area:AI-agent-guardrails",
      "to": "domain:security",
      "kind": "applies_to",
      "attributes": {
        "confidence": "secondary"
      }
    },
    {
      "from": "skill-area:AI-agent-guardrails",
      "to": "skill-area:AI-safety-alignment",
      "kind": "prerequisite_for_learning",
      "attributes": {
        "strength": "recommended"
      }
    },
    {
      "from": "skill-area:AI-agent-guardrails",
      "to": "skill-area:AI-agent-evaluation",
      "kind": "prerequisite_for_learning",
      "attributes": {
        "strength": "helpful"
      }
    }
  ],
  "incomingEdges": []
}