Reason-State-Queue-Resume

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 priority queue pops highest priority first. Start with A:1, B:3, C:2. Pop once, insert D:2, then pop twice. Which jobs were popped?

{
  "metrics": {
    "accuracy": {
      "checks": [
        {
          "contains": [
            "B"
          ]
        },
        {
          "contains": [
            "C"
          ]
        },
        {
          "contains": [
            "D"
          ]
        }
      ]
    },
    "evidence": {
      "checks": [
        {
          "contains": [
            "priority"
          ]
        },
        {
          "contains": [
            "insert D"
          ]
        },
        {
          "contains": [
            "pop"
          ]
        }
      ]
    },
    "constraint": {
      "checks": [
        {
          "contains": [
            "stateful"
          ]
        },
        {
          "contains": [
            "tie"
          ]
        },
        {
          "contains": [
            "highest first"
          ]
        }
      ]
    }
  }
}