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 4 short bullet points:
A platform team improved release safety by introducing load testing, change approvals, and dead-letter queues across critical services. They documented ownership, ran weekly reliability reviews, and tracked defect escape rate to verify impact over time.

{
  "expected_contains": [
    "load testing",
    "change approvals",
    "dead-letter queues",
    "defect escape rate"
  ]
}