Reason-Constraint-Consistency-Latency

The model receives the prompt (and optional system message). The run uses scorer rubric_json_metrics with the JSON configuration below. Pass/fail and partial credit are determined entirely by that scorer against the model output; no human grading.

Return JSON only with keys answer, evidence, constraints. A product requires read-your-writes consistency and p99 under 20 ms across continents. Explain the tension and what tradeoff must be made.

{
  "metrics": {
    "accuracy": {
      "checks": [
        {
          "contains": [
            "read-your-writes"
          ]
        },
        {
          "contains": [
            "p99 under 20 ms"
          ]
        },
        {
          "contains": [
            "continents"
          ]
        }
      ]
    },
    "evidence": {
      "checks": [
        {
          "contains": [
            "consistency"
          ]
        },
        {
          "contains": [
            "latency"
          ]
        },
        {
          "contains": [
            "tradeoff"
          ]
        }
      ]
    },
    "constraint": {
      "checks": [
        {
          "contains": [
            "regional routing"
          ]
        },
        {
          "contains": [
            "replication"
          ]
        },
        {
          "contains": [
            "constraint"
          ]
        }
      ]
    }
  }
}