Summary

The model receives the prompt (and optional system message). The run uses scorer contains_all 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 exactly 3 short bullet points: Incident communication became clearer by defining severity levels, status page templates, and stakeholder update cadence. Postmortems captured owners and due dates.

{
  "expected_contains": [
    "severity",
    "status page",
    "postmortems"
  ]
}