{
  "model_type": "cognitive-engine",
  "loop": ["perceive", "reason", "act", "reflect"],
  "learning_signal": "human_feedback",
  "explainable": true
}