id: '147'
narrative_anchor_date: '2028-01-01'
test: For experiment EXP-2023-02-onboarding-trigger, fill only the missing learning-carryforward field.
  State the second cohort's result, the decision it caused, who made that decision, the planned rollout
  date and percentage, who was preparing it, and whose checks could stop it. Do not change status, result,
  or landed date.
load_bearing_facts:
- 352
- 353
expected_tool_calls:
- update_experiment
grade:
  type: tool_trace
  config:
    check_version: 2
    today: '2028-01-01'
    semantic_judge_version: 2
    assertions:
    - type: field_equals
      tool: update_experiment
      action_id: backfill_learning
      path: result.ok
      value: true
      check_id: riley_147_00
    - type: field_equals
      tool: update_experiment
      action_id: backfill_learning
      path: args.experiment_id
      value: EXP-2023-02-onboarding-trigger
      check_id: riley_147_01
    - type: field_absent_or_empty
      tool: update_experiment
      action_id: backfill_learning
      path: args.status
      check_id: riley_147_02
    - type: field_absent_or_empty
      tool: update_experiment
      action_id: backfill_learning
      path: args.result
      check_id: riley_147_03
    - type: field_absent_or_empty
      tool: update_experiment
      action_id: backfill_learning
      path: args.landed_at
      check_id: riley_147_04
    - type: field_llm_judge
      tool: update_experiment
      action_id: backfill_learning
      path: args
      criterion: Riley called the experiment successful after the second cohort cleared. Riley and Ines
        planned a 100% rollout for April 10, 2023 unless Owen's pull showed sample-ratio mismatch, signup-source
        mix problems, or unusual country or device skew.
      check_id: riley_147_05
mock_state:
  experiments:
  - experiment_id: EXP-2023-02-onboarding-trigger
    name: EXP-2023-02-onboarding-trigger
    hypothesis: week-1 retention moves at least +2 pts if the email lands when they have actually crossed
      the first-use threshold, not while they are still half-dead in setup.
    metric_of_record: week-1 retention
    cohort_scope: mid-seg only
    flag_id: lifecycle_onboarding_trigger_v2
    status: shipped
    designed_at: '2023-03-01'
    created_at: '2023-03-01T14:00:00-06:00'
    updated_at: '2023-04-10T10:00:00-05:00'
    landed_at: '2023-04-10'