Test 095

Jan 1, 2028 / 2 facts

Request

Create an "Infra release review" event for January 12, 2028 from 2:00 to 2:45 PM Eastern with infra-release@example.com and product-eng@example.com, and add the appropriate internal review surface to the agenda.

Required memory

Fact 92

Theo approved Lantern’s narrow internal-adoption step, allowing its use in selected release reviews and incident follow-ups for Infra and Product Engineering; Hema backed the decision because the access and content boundaries remained explicit.

Source evidence (1)

000708Nov 7, 2023 / 11:40 UTC-05:00

The Lantern decision meeting ended. Theo approved the narrow internal-adoption step, and Hema backed it because the access and content boundaries stayed explicit. Lantern may now be used in selected release reviews and incident follow-ups for Infra and Product Engineering. Access stays bounded. The surface remains deploy movement, provenance-backed ownership, and incident-load summaries. It does not expose raw incident bodies, infer readiness, become customer-facing, or become company-wide. Create a short Eng doc immediately so the selected-room approval doesn't turn into hallway scope creep.

Message 000708 in history

Fact 93

Lantern access is limited to selected internal operating rooms and existing responsible participants.

Source evidence (1)

000709Nov 7, 2023 / 11:40 UTC-05:00

# Lantern narrow adoption — Nov 7 decision ## Decision Theo approved a narrow internal-adoption step for Lantern, with Hema's backing. Lantern may be used in selected release reviews and incident follow-ups for Infra and Product Engineering. ## Access boundary Access stays bounded to the selected internal operating rooms and existing responsible participants. This is not a company-wide rollout. ## Surface boundary The adopted surface shows: - deploy movement, - provenance-backed ownership, - incident-load summaries, - empty states when the source data is missing or not permissioned. The adopted surface does not show raw incident bodies and does not make readiness judgments. ## Ownership/provenance rule Ownership cards must be backed by recorded owner-map provenance. Missing owner data should be fixed at the owner-map source, not patched manually in the UI. ## Non-goals - Not customer-facing. - Not company-wide. - Not a manager-readiness dashboard. - No raw incident bodies. - No release-readiness scoring. - No manual owner-card overrides.

Message 000709 in history

Expected tool calls

create_calendar_event

Grading

1. field_equals / create_calendar_event

{
  "type": "field_equals",
  "tool": "create_calendar_event",
  "action_id": "alex_095_create_calendar_event",
  "path": "result.ok",
  "value": true,
  "check_id": "alex_095_00"
}

2. field_equals / create_calendar_event

{
  "type": "field_equals",
  "path": "result.event.title",
  "value": "Infra release review",
  "tool": "create_calendar_event",
  "action_id": "alex_095_create_calendar_event",
  "check_id": "alex_095_01"
}

3. datetime_absolute_eq / create_calendar_event

{
  "type": "datetime_absolute_eq",
  "path": "result.event.start",
  "value": "2028-01-12T14:00:00-05:00",
  "tool": "create_calendar_event",
  "action_id": "alex_095_create_calendar_event",
  "check_id": "alex_095_02"
}

4. datetime_absolute_eq / create_calendar_event

{
  "type": "datetime_absolute_eq",
  "path": "result.event.end",
  "value": "2028-01-12T14:45:00-05:00",
  "tool": "create_calendar_event",
  "action_id": "alex_095_create_calendar_event",
  "check_id": "alex_095_03"
}

5. field_list_includes / create_calendar_event

{
  "type": "field_list_includes",
  "path": "result.event.attendees",
  "values": [
    "infra-release@example.com",
    "product-eng@example.com"
  ],
  "tool": "create_calendar_event",
  "action_id": "alex_095_create_calendar_event",
  "check_id": "alex_095_04"
}

6. field_llm_judge / create_calendar_event

{
  "type": "field_llm_judge",
  "path": "result.event.body",
  "criterion": "The agenda names Lantern as the internal review surface for this selected Infra and Product Engineering release review and does not imply general or unrestricted availability.",
  "tool": "create_calendar_event",
  "action_id": "alex_095_create_calendar_event",
  "check_id": "alex_095_05"
}

Complete grading specification

{
  "type": "tool_trace",
  "config": {
    "check_version": 2,
    "today": "2028-01-01",
    "assertions": [
      {
        "type": "field_equals",
        "tool": "create_calendar_event",
        "action_id": "alex_095_create_calendar_event",
        "path": "result.ok",
        "value": true,
        "check_id": "alex_095_00"
      },
      {
        "type": "field_equals",
        "path": "result.event.title",
        "value": "Infra release review",
        "tool": "create_calendar_event",
        "action_id": "alex_095_create_calendar_event",
        "check_id": "alex_095_01"
      },
      {
        "type": "datetime_absolute_eq",
        "path": "result.event.start",
        "value": "2028-01-12T14:00:00-05:00",
        "tool": "create_calendar_event",
        "action_id": "alex_095_create_calendar_event",
        "check_id": "alex_095_02"
      },
      {
        "type": "datetime_absolute_eq",
        "path": "result.event.end",
        "value": "2028-01-12T14:45:00-05:00",
        "tool": "create_calendar_event",
        "action_id": "alex_095_create_calendar_event",
        "check_id": "alex_095_03"
      },
      {
        "type": "field_list_includes",
        "path": "result.event.attendees",
        "values": [
          "infra-release@example.com",
          "product-eng@example.com"
        ],
        "tool": "create_calendar_event",
        "action_id": "alex_095_create_calendar_event",
        "check_id": "alex_095_04"
      },
      {
        "type": "field_llm_judge",
        "path": "result.event.body",
        "criterion": "The agenda names Lantern as the internal review surface for this selected Infra and Product Engineering release review and does not imply general or unrestricted availability.",
        "tool": "create_calendar_event",
        "action_id": "alex_095_create_calendar_event",
        "check_id": "alex_095_05"
      }
    ]
  }
}

App state

Download JSON

Source file

tests/alex/095.yaml

SHA-256: f9104ef60fc248235dced42bce8feb42fb400a1d38ab74f3788d0959c44c7d90