Cost-Complex-Debounce-Replay

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 benchmark, title, code, done. Set benchmark to "cost", title to "Simulate a debounced call schedule", and done to true. Given call timestamps and waitMs, return execution timestamps for trailing debounce.

{
  "metrics": {
    "strict": {
      "checks": [
        {
          "contains": [
            "cost"
          ]
        },
        {
          "contains": [
            "done"
          ]
        },
        {
          "contains": [
            "debounce"
          ]
        },
        {
          "contains": [
            "trailing"
          ]
        },
        {
          "contains": [
            "timestamps"
          ]
        }
      ]
    }
  }
}