Refactor-Date-Format-Helper

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 plan, refactor, tests. Three modules repeat slightly different date formatting snippets for audit logs. Extract shared behavior while keeping caller-specific labels.

{
  "metrics": {
    "intent": {
      "checks": [
        {
          "contains": [
            "duplicate date formatting"
          ]
        },
        {
          "contains": [
            "shared helper"
          ]
        },
        {
          "contains": [
            "audit logs"
          ]
        }
      ]
    },
    "visible": {
      "checks": [
        {
          "contains": [
            "three modules"
          ]
        },
        {
          "contains": [
            "caller-specific labels"
          ]
        },
        {
          "contains": [
            "same format"
          ]
        }
      ]
    },
    "hidden": {
      "checks": [
        {
          "contains": [
            "unit tests"
          ]
        },
        {
          "contains": [
            "timezone"
          ]
        },
        {
          "contains": [
            "no behavior change"
          ]
        }
      ]
    }
  }
}