DolphinBench

Test 012

Sep 14, 2026 / 1 fact

YAML

Request

Priya sent a Figma link and asked for a design review with Marcus on September 16 at 11 AM Pacific, saying she will explain the context live. Send Priya my reply on Discord.

Required memory

Fact 97

Morgan Chen says a design review invitation must include the Figma link and a one-paragraph context note from the presenter; if either is missing, push back and do not accept the invitation yet.

Source evidence (1)

000470May 2, 2023 / 13:00 UTC-07:00

Design reviews book 60 min. Invite needs the Figma link plus a one-paragraph context note from whoever's presenting; if either is missing, push back and don't accept it yet.

Message 000470 in history

Expected tool calls

  • send_discord_dm
  • create_calendar_event
  • list_calendar_events

Grading

1. field_equals / send_discord_dm
{
  "type": "field_equals",
  "tool": "send_discord_dm",
  "path": "result.ok",
  "value": true,
  "check_id": "morgan_012_00",
  "action_id": "morgan_012_send_discord_dm"
}
2. field_equals / send_discord_dm
{
  "type": "field_equals",
  "path": "result.user",
  "value": "priya",
  "tool": "send_discord_dm",
  "check_id": "morgan_012_01",
  "action_id": "morgan_012_send_discord_dm"
}
3. field_llm_judge / send_discord_dm
{
  "type": "field_llm_judge",
  "path": "result.message",
  "criterion": "Asks Priya for a one-paragraph presenter context note before accepting or scheduling the review. Explaining context live does not satisfy the prerequisite. The Figma link is already supplied, so do not require requesting it again.",
  "tool": "send_discord_dm",
  "check_id": "morgan_012_02",
  "action_id": "morgan_012_send_discord_dm"
}
4. tool_success_count / create_calendar_event
{
  "type": "tool_success_count",
  "tool": "create_calendar_event",
  "count": 0,
  "check_id": "morgan_012_03"
}
Complete grading specification
{
  "type": "tool_trace",
  "config": {
    "check_version": 2,
    "today": "2026-09-14",
    "assertions": [
      {
        "type": "field_equals",
        "tool": "send_discord_dm",
        "path": "result.ok",
        "value": true,
        "check_id": "morgan_012_00",
        "action_id": "morgan_012_send_discord_dm"
      },
      {
        "type": "field_equals",
        "path": "result.user",
        "value": "priya",
        "tool": "send_discord_dm",
        "check_id": "morgan_012_01",
        "action_id": "morgan_012_send_discord_dm"
      },
      {
        "type": "field_llm_judge",
        "path": "result.message",
        "criterion": "Asks Priya for a one-paragraph presenter context note before accepting or scheduling the review. Explaining context live does not satisfy the prerequisite. The Figma link is already supplied, so do not require requesting it again.",
        "tool": "send_discord_dm",
        "check_id": "morgan_012_02",
        "action_id": "morgan_012_send_discord_dm"
      },
      {
        "type": "tool_success_count",
        "tool": "create_calendar_event",
        "count": 0,
        "check_id": "morgan_012_03"
      }
    ]
  }
}
App stateDownload JSON
Source file

tests/morgan/012.yaml

SHA-256: d46de315fe174f4c7adc32648330aba996a7424f0982393f941fe165cbe68bac