DolphinBench

Test 118

Jan 1, 2028 / 2 facts

YAML

Request

For our five-year internal retrospective, create a citation-ready archive note about the onboarding-trigger result as presented at the formal May 12, 2023 company all-hands. Capture the audience scale, how the result was framed, and the contributors explicitly credited in that presentation. Include a brief source note based on the contemporaneous records, and keep the recognition specific to this all-hands setting rather than presenting it as a universal credit rule.

Required memory

Fact 378

Riley Tanaka delivered the full 15-minute presentation and 10-minute Q&A on mid-segment churn to about 140 people at the May 12 all-hands.

Source evidence (1)

000237May 12, 2023 / 14:00 UTC-05:00

all-hands went well. I did the full 15 + 10 on the mid-segment churn thing to the whole ~140, kept it on the method instead of making it a drama deck. the sequence landed: cohort cut, enumerate hypotheses, rule-out cadence, then A/B, then wait for the stabilization cohort before acting like you know anything. Marcus Vail asked a nice question, very much in the room with me not testing me. Tomas pushed on whether it generalizes to other cohorts and I just said yeah, we should use it elsewhere. also got Owen Pham + Ines Kovac in there by name. pretty sure this doc is gonna become one of those internal refs people keep passing around -- which is correct.

Message 000237 in history

Fact 379

Riley Tanaka explicitly credited Owen Pham and Ines Kovac by name during the May 12 all-hands presentation.

Source evidence (1)

000237May 12, 2023 / 14:00 UTC-05:00

all-hands went well. I did the full 15 + 10 on the mid-segment churn thing to the whole ~140, kept it on the method instead of making it a drama deck. the sequence landed: cohort cut, enumerate hypotheses, rule-out cadence, then A/B, then wait for the stabilization cohort before acting like you know anything. Marcus Vail asked a nice question, very much in the room with me not testing me. Tomas pushed on whether it generalizes to other cohorts and I just said yeah, we should use it elsewhere. also got Owen Pham + Ines Kovac in there by name. pretty sure this doc is gonna become one of those internal refs people keep passing around -- which is correct.

Message 000237 in history

Expected tool calls

  • create_doc

Grading

1. field_equals / create_doc
{
  "type": "field_equals",
  "tool": "create_doc",
  "action_id": "create_archive_note",
  "path": "result.ok",
  "value": true,
  "check_id": "riley_118_00"
}
2. field_llm_judge / create_doc
{
  "type": "field_llm_judge",
  "tool": "create_doc",
  "action_id": "create_archive_note",
  "path": "args",
  "criterion": "The title clearly identifies the document as an archive or retrospective note about the May 12, 2023 all-hands presentation.\n\nThe body accurately says that the formal May 12, 2023 company all-hands reached the full-company audience of approximately 140 people and presented the onboarding-trigger result within a broader investigation method or framework, with a method-focused rather than dramatic or postmortem framing. Missing or contradictory audience or framing information must fail.\n\nThe body states that Owen Pham and Ines Kovac were explicitly credited by name in the May 12 all-hands presentation. Both full names are required, and substitution, omission, or contradiction must fail.\n\nThe body makes clear that these contributor credits describe the formal May 12 all-hands presentation setting and does not turn them into a universal credit rule for every communication or use of the framework.\n\nThe body includes a brief citation-ready source note grounding the factual account in one or more contemporaneous May 12, 2023 records, without inventing external sources.",
  "check_id": "riley_118_01"
}
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_archive_note",
        "path": "result.ok",
        "value": true,
        "check_id": "riley_118_00"
      },
      {
        "type": "field_llm_judge",
        "tool": "create_doc",
        "action_id": "create_archive_note",
        "path": "args",
        "criterion": "The title clearly identifies the document as an archive or retrospective note about the May 12, 2023 all-hands presentation.\n\nThe body accurately says that the formal May 12, 2023 company all-hands reached the full-company audience of approximately 140 people and presented the onboarding-trigger result within a broader investigation method or framework, with a method-focused rather than dramatic or postmortem framing. Missing or contradictory audience or framing information must fail.\n\nThe body states that Owen Pham and Ines Kovac were explicitly credited by name in the May 12 all-hands presentation. Both full names are required, and substitution, omission, or contradiction must fail.\n\nThe body makes clear that these contributor credits describe the formal May 12 all-hands presentation setting and does not turn them into a universal credit rule for every communication or use of the framework.\n\nThe body includes a brief citation-ready source note grounding the factual account in one or more contemporaneous May 12, 2023 records, without inventing external sources.",
        "check_id": "riley_118_01"
      }
    ]
  }
}
App stateDownload JSON
Source file

tests/riley/118.yaml

SHA-256: 418c54ab32fafb4fde9d87198c712379c7bcf3e522d92c0ff8ecae9941e438be