{
  "context": "Tests whether the agent validates resume_token format for both packets — catching a token that is too short and one that contains disallowed characters/spaces — then classifies and reports appropriately.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "Token check performed",
      "description": "Output includes an explicit token validation or resume_token check for both packets",
      "max_score": 10
    },
    {
      "name": "Packet A token rejected",
      "description": "Packet A's token ('ab!') is identified as invalid (too short, invalid characters, or both)",
      "max_score": 15
    },
    {
      "name": "Packet B token rejected",
      "description": "Packet B's token is identified as invalid (contains spaces and/or disallowed special characters)",
      "max_score": 15
    },
    {
      "name": "Format requirements stated",
      "description": "The review mentions at least one token format constraint: minimum 8 characters, maximum 128 characters, or that only alphanumeric/underscore/dash characters are allowed",
      "max_score": 12
    },
    {
      "name": "Non-clean classification for both",
      "description": "Neither packet is classified as CLEAN — both receive OPERATIONAL or CRITICAL",
      "max_score": 12
    },
    {
      "name": "Recovery includes new token",
      "description": "Recovery steps for at least one packet include generating or providing a new valid resume token",
      "max_score": 12
    },
    {
      "name": "Per-check summary both packets",
      "description": "Output includes a per-check breakdown (schema, freshness, token, replay or similar) for each of the two packets",
      "max_score": 12
    },
    {
      "name": "Escalation for both packets",
      "description": "Output includes escalation recommendations for both Packet A and Packet B",
      "max_score": 12
    }
  ]
}