Advanced OpenClaw Quiz

{
  "type": "quiz",
  "passThreshold": 0.7,
  "questions": [
    {
      "id": "q1",
      "text": "When should you split work across multiple agents rather than using a single agent?",
      "points": 10,
      "options": [
        {"id": "a", "text": "Always, because multiple agents are more professional"},
        {"id": "b", "text": "When tasks are independent, run on different schedules, or need different models or security boundaries"},
        {"id": "c", "text": "Only when you have more than 100 cron jobs"},
        {"id": "d", "text": "Never, because a single agent is always more efficient"}
      ],
      "correctAnswers": ["b"]
    },
    {
      "id": "q2",
      "text": "In the Monitor and Actor agent pattern, what is the purpose of using a cheap local model for the watchdog agent?",
      "points": 10,
      "options": [
        {"id": "a", "text": "Local models are always more accurate for monitoring tasks"},
        {"id": "b", "text": "It keeps costs at zero for frequent checks that usually find nothing, saving the cloud model for when action is needed"},
        {"id": "c", "text": "Local models can run without electricity"},
        {"id": "d", "text": "The watchdog agent does not need AI capabilities at all"}
      ],
      "correctAnswers": ["b"]
    },
    {
      "id": "q3",
      "text": "What is a fallback chain in OpenClaw?",
      "points": 1