{
  "storage_consent": false,
  "allow_public_summary": false,
  "redacted_attestation": false,
  "schema_version": "risk-batch-v1",
  "cohort": {
    "model": "your-model",
    "model_revision": "your-model-revision",
    "task": "support_credit",
    "workflow_version": "your-workflow-v1",
    "policy_version": "your-policy-v1",
    "memory": true,
    "critic": false,
    "guidance_version": "your-memory-version",
    "critic_version": "none",
    "guard_version": "none"
  },
  "dataset_id": "replace-with-unique-evaluation-id",
  "source": "observed",
  "sampling_design": "unspecified",
  "sampling_description": "Describe how attempted tasks were selected, including successful and failed tasks.",
  "failure_definition": "Define the independent rule used to classify a task as failed.",
  "loss_measure": "planning",
  "currency": "USD",
  "loss_cap_cents": 250000,
  "episodes": [
    {
      "episode_id": "replace-task-1",
      "reviewed": false,
      "review_ref": null,
      "task_failure": null,
      "loss_cents": null
    }
  ]
}
