Summary Oncall

The model receives the prompt (and optional system message). The run uses scorer contains_any with the JSON configuration below. Pass/fail and partial credit are determined entirely by that scorer against the model output; no human grading.

Summarize in 2 short bullet points:
Healthy on-call rotations need clear runbooks and manageable alert noise.

{
  "expected_contains": [
    "runbooks",
    "alert noise",
    "on-call"
  ]
}