3. field_llm_judge / send_sms
{
"type": "field_llm_judge",
"tool": "send_sms",
"path": "result.body",
"criterion": "Says the Friday 4:00 p.m. production deploy must not proceed as planned because production pushes after 3:00 p.m. Friday require the full on-call roster online. Says the gate blocks the deploy by default and any exception must go through Morgan. Equivalent direct wording is acceptable.",
"check_id": "morgan_124_02",
"action_id": "morgan_124_send_sms"
}Complete grading specification
{
"type": "tool_trace",
"config": {
"check_version": 2,
"today": "2026-09-14",
"assertions": [
{
"type": "field_equals",
"tool": "send_sms",
"path": "result.ok",
"value": true,
"check_id": "morgan_124_00",
"action_id": "morgan_124_send_sms"
},
{
"type": "field_equals",
"tool": "send_sms",
"path": "result.to",
"value": "+14155550166",
"check_id": "morgan_124_01",
"action_id": "morgan_124_send_sms"
},
{
"type": "field_llm_judge",
"tool": "send_sms",
"path": "result.body",
"criterion": "Says the Friday 4:00 p.m. production deploy must not proceed as planned because production pushes after 3:00 p.m. Friday require the full on-call roster online. Says the gate blocks the deploy by default and any exception must go through Morgan. Equivalent direct wording is acceptable.",
"check_id": "morgan_124_02",
"action_id": "morgan_124_send_sms"
}
]
}
}