python-sdk/pkg/hanzo-kai/tests/conformance/ids.json

{
  "rule": "Question ids are never visible to the model: renaming every id gives bit-identical answers.",
  "cases": [
    {
      "path": [
        "native",
        "compat"
      ],
      "send": [
        {
          "state": "I was charged twice for my March invoice. Please refund the duplicate.",
          "questions": {
            "billing": {
              "type": "noul",
              "instructions": "Is this about billing?",
              "criteria": {
                "true": "it concerns charges, invoices or refunds",
                "false": "it concerns something else"
              }
            },
            "team": {
              "type": "choice",
              "instructions": "Which team should handle this?",
              "criteria": {
                "billing": "charges, invoices, refunds",
                "tech": "bugs and outages",
                "sales": "plans and upgrades"
              }
            },
            "urgency": {
              "type": "score",
              "instructions": "How urgent is this?",
              "criteria": [
                "can wait",
                "this week",
                "today",
                "right now"
              ]
            }
          }
        },
        {
          "state": "I was charged twice for my March invoice. Please refund the duplicate.",
          "questions": {
            "a": {
              "type": "noul",
              "instructions": "Is this about billing?",
              "criteria": {
                "true": "it concerns charges, invoices or refunds",
                "false": "it concerns something else"
              }
            },
            "b": {
              "type": "choice",
              "instructions": "Which team should handle this?",
              "criteria": {
                "billing": "charges, invoices, refunds",
                "tech": "bugs and outages",
                "sales": "plans and upgrades"
              }
            },
            "c": {
              "type": "score",
              "instructions": "How urgent is this?",
              "criteria": [
                "can wait",
                "this week",
                "today",
                "right now"
              ]
            }
          }
        }
      ],
      "expect": {
        "status": 200,
        "checks": [
          "same_answers"
        ]
      }
    }
  ]
}