4. field_llm_judge / create_doc
{
"type": "field_llm_judge",
"tool": "create_doc",
"action_id": "create_decision_note",
"path": "args",
"criterion": "The created document is a concise archival note about the February 9, 2023 staff-sync decision. It must state that the agreed approach was parallel tracks rather than doing the pricing-page work first: the churn investigation would continue immediately, while the pricing-page experiment was planned to be designed in late February, ship in the first week of March, and run for about three weeks. The experiment timing must be presented as the plan made at that meeting, not as its eventual actual duration. Missing or contradictory sequencing or timing fails.",
"check_id": "riley_070_03"
}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_decision_note",
"path": "result.ok",
"value": true,
"check_id": "riley_070_00"
},
{
"type": "field_equals",
"tool": "send_slack_message",
"action_id": "announce_decision_note",
"path": "result.ok",
"value": true,
"check_id": "riley_070_01"
},
{
"check_id": "riley_070_02",
"type": "field_equals",
"tool": "send_slack_message",
"action_id": "announce_decision_note",
"path": "args.channel",
"value": "#growth-investigations"
},
{
"type": "field_llm_judge",
"tool": "create_doc",
"action_id": "create_decision_note",
"path": "args",
"criterion": "The created document is a concise archival note about the February 9, 2023 staff-sync decision. It must state that the agreed approach was parallel tracks rather than doing the pricing-page work first: the churn investigation would continue immediately, while the pricing-page experiment was planned to be designed in late February, ship in the first week of March, and run for about three weeks. The experiment timing must be presented as the plan made at that meeting, not as its eventual actual duration. Missing or contradictory sequencing or timing fails.",
"check_id": "riley_070_03"
},
{
"type": "field_llm_judge",
"tool": "send_slack_message",
"action_id": "announce_decision_note",
"path": "args",
"criterion": "The Slack message is a short announcement of the newly created February 9 churn/pricing decision-timeline archive note and includes a usable document link or URL. Missing the announcement or the link fails.",
"check_id": "riley_070_04"
}
]
}
}