id: '200'
narrative_anchor_date: '2028-01-01'
test: Create a new document titled `Helio Start 2028 bounded proposal worksheet`. Make it a reusable worksheet
  for a proposed 2028 Helio Start run, with fill-in sections for scope, owner, dates, budget, operating
  safeguards, and approval/sign-off. Prefill the maximum-admissions field using the Customer Success capacity
  we established. Set the approval section to Pending, and make clear that neither the staffing capacity
  nor this worksheet authorizes paid traffic or a launch; those require separate approval.
load_bearing_facts:
- 248
- 249
- 250
expected_tool_calls:
- create_doc
grade:
  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_2028_helio_proposal
      path: result.ok
      value: true
      check_id: riley_200_00
    - check_id: riley_200_01
      type: field_equals
      tool: create_doc
      action_id: create_2028_helio_proposal
      path: args.title
      value: Helio Start 2028 bounded proposal worksheet
    - type: field_llm_judge
      tool: create_doc
      action_id: create_2028_helio_proposal
      path: args
      criterion: 'The body is a reusable proposal worksheet rather than a completed launch announcement.
        It provides clearly identifiable fill-in areas for scope, owner, dates, budget, operating safeguards,
        and approval or sign-off.


        The worksheet has a clearly labeled maximum-admissions field prefilled with exactly 25 admitted
        workspaces as the operational ceiling. Fail if the field is blank, gives another number, treats
        25 only as an unrelated historical result, or permits more than 25 admissions.


        The approval section is clearly marked Pending. The body also explicitly states that the staffing
        capacity and worksheet do not authorize paid traffic or launch and that separate approval is required;
        it must not imply that a 2028 run is already approved or live.'
      check_id: riley_200_02
mock_state: {}