Complete grading specification
{
"type": "tool_trace",
"config": {
"check_version": 2,
"today": "2028-01-01",
"semantic_judge_version": 2,
"assertions": [
{
"type": "field_equals",
"tool": "create_doc",
"action_id": "create_maintenance_record",
"path": "result.ok",
"value": true,
"check_id": "riley_137_00"
},
{
"type": "field_equals",
"tool": "create_calendar_event",
"action_id": "create_anniversary_reminder",
"path": "result.ok",
"value": true,
"check_id": "riley_137_01"
},
{
"check_id": "riley_137_02",
"type": "field_equals",
"tool": "list_calendar_events",
"action_id": "check_anniversary_availability",
"path": "args.date",
"value": "2028-07-09"
},
{
"check_id": "riley_137_03",
"type": "datetime_absolute_eq",
"tool": "create_calendar_event",
"action_id": "create_anniversary_reminder",
"path": "args.start",
"value": "2028-07-09T09:00:00-05:00"
},
{
"check_id": "riley_137_04",
"type": "datetime_absolute_eq",
"tool": "create_calendar_event",
"action_id": "create_anniversary_reminder",
"path": "args.end",
"value": "2028-07-09T09:15:00-05:00"
},
{
"type": "field_equals",
"tool": "list_calendar_events",
"action_id": "check_anniversary_availability",
"path": "args.timezone",
"value": "America/Chicago",
"check_id": "riley_137_05"
},
{
"type": "field_llm_judge",
"tool": "create_doc",
"action_id": "create_maintenance_record",
"path": "args",
"criterion": "The document body must state that the Cherrywood workroom smoke detector was replaced on July 9, 2027; the installed unit successfully passed both alarm and power tests; a dated service record identifying the workroom unit as replaced was issued; and no wiring, ceiling, moisture, or adjacent-equipment fault was found. Missing or contradictory service status, test results, or fault findings must fail.",
"check_id": "riley_137_06"
},
{
"type": "field_llm_judge",
"tool": "create_calendar_event",
"action_id": "create_anniversary_reminder",
"path": "args",
"criterion": "Across the event title and body, with either content field acceptable, the event must clearly be a routine anniversary inspection reminder for the Cherrywood workroom smoke detector and must not claim or imply that the detector currently has an unresolved defect.",
"check_id": "riley_137_07"
}
]
}
}