Debug-Lru-Eviction-V2

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 diagnosis, fix, tests. An LRU cache refactor updates values without refreshing access order, so recently written keys are evicted. Diagnose and fix the JavaScript logic.

{
  "metrics": {
    "repro": {
      "checks": [
        {
          "contains": [
            "set existing"
          ]
        },
        {
          "contains": [
            "access order"
          ]
        },
        {
          "contains": [
            "evict"
          ]
        },
        {
          "contains": [
            "recent"
          ]
        }
      ]
    },
    "hidden": {
      "checks": [
        {
          "contains": [
            "delete"
          ]
        },
        {
          "contains": [
            "reinsert"
          ]
        },
        {
          "contains": [
            "map order"
          ]
        }
      ]
    },
    "diagnose": {
      "checks": [
        {
          "contains": [
            "LRU"
          ]
        },
        {
          "contains": [
            "update"
          ]
        },
        {
          "contains": [
            "stale order"
          ]
        }
      ]
    }
  }
}