Evidence status: author-simulated teacher, learner, parent/carer and potential funder desk roles only. No independent person/classroom took part; no actual place/child/speaker/care record was collected. No access, teaching quality, learning, interest, consent or market result is claimed.
Teacher role: I traced folder marks through the two distinct source diagrams and kept the original records. The whole later record accounts for all four even though the shelf has fewer. C supplies a step, D one person's reason, E a proposed message and F a wording check. None fills B's unknown mover/cause or proves use. I reviewed all ten practical scripts: feature desk, identity ledger, request flap, named reason card, fresh check, planned cue, repair station, permitted care use, desk tour and second fresh check. Each contains six timed stages. The eight actual branches include named inputs, exact short questions, source/view/time records, real care procedure, useful permitted task/cue and action-help record, with a full paper fallback. Their evidence can support K03 elements without treating a template as completion. Local language/pacing, safe task, source selection and curriculum judgement remain educator gates.
Learner role: I walked through print/point, exact oral and adult-directed broad-piece routes on the desk. An individual mark's location can be followed without testing number reading; an empty container does not erase an item listed elsewhere. I followed the request-flap, planned-cue and source-correction instructions in a desk simulation; I did not physically construct a paper/tactile desk. In the fresh second case I could keep the reported actor for one item separate from the other item's unknown actor/cause. I could choose explanation instead of touching/performing care and still show source reasoning. These are author interpretations, not learner understanding/interest/independence evidence. Raised layouts and AAC choices were specified/walked through verbally; none was physically child-tested.
Parent/carer role: I checked ten free, optional invented-paper bridges with school equivalents and no return/score, home map, address, real-property movement, device, family interview or cultural disclosure. Pass has no assessment cost. Real-source work is preflighted school work; it doesn't demand private home evidence. Home language can carry reasoning, while supplied translations would need fluent human/local review. No parent/carer reviewed the pack.
Potential funder role: I checked authored completeness: ten 25-minute sessions, thirty routes, twenty worked everyday-domain variations, ten bridges, two fresh skill checks/separate key, nine original A4 pairs/alternatives, integrated actual K03 pathway and standard care record, precise official source rows, rights and complete hashes. These establish a reproducible draft, not teacher time saved, learning gain, completed knowledge, engagement, market demand, universal access or commercial suitability. K01/K02/K04 remain separate authentic-source gates.
Residual critique: matching unique identity across records is demanding; use one broad mark at a time and exact source wording. The mark isn't a numeral-reading test. Actual views may be partial or identity matching unreliable: preserve that limit and use paper sources for this skill instead. Picture-message making and small tasks may use the seven-minute inquiry; pre-make frames and offer complete adult-directed construction. Care benefit and independent action must not be inferred from a finished cue/task. Individual print, screen reader, tactile, AAC, language and pacing review remain required; PDFs are untagged. A real local explanation can support care understanding with no compulsory action; missing connection/meaning/features remain open.
Reproduction: python print/generate_print.py --regenerate deliberately rewrites only nine named SVG/PDF pairs and alternatives. python verify_pack.py --manifest-json prints a proposed complete receipt; author deliberately saves it separately. Default python verify_pack.py is offline/read-only. Python 3/Pillow/DejaVu, librsvg, pdfinfo, pdftotext and pdffonts are required. Fixed PDF build clock is reproduction metadata, not an actual source/event date.
Final technical receipt · 30 September 2026
- Offline read-only audit: PASS — nine exact workbook XML rows/imported records/crosswalk and QCAA pins; ten 25-minute scripts with eight integrated actual-source/place branches; conditional K03 with no outcomes observed; 30 switchable routes; 20 worked variations; 10 bridges; two fresh checks/separate key; nine A4 pairs, exact alternatives, measured text bounds, searchable embedded fonts and keyed item identity/location geometry; 57 local links; 37 hashed files (38 total with manifest).
- Reproduction: second deliberate regeneration produced byte-identical nine SVGs, nine PDFs and complete alternatives file: 19/19 artifact hashes matched. Fixed build clock is metadata only.
- Negative checks: 15/15 in-memory mutations rejected: workbook digest; exact official description; K01 hold; blanket K03 hold; invented actual outcome; daily timing; missing integrated actual branch; missing useful real-care task; missing no-actual-evidence boundary; fresh speaker leakage; learner-key leakage; alternative text drift; printed/source quote drift; item identity/place geometry; stale receipt. No bad fixture was written. Normal audit and negative exercises preserved every pack file's bytes, size and modification time; original inputs were restored in memory.
- Rendered page inspection: all nine PDFs rendered and viewed together. First inspection found an ornamental speech outline touching Pia's source panel and a crowded bottom spacing on Puzzle Point. Those were corrected; both pages were re-rendered and inspected. Final diagrams match item marks/locations, text remains within panels and page bounds, and footer/source/blank-record fields are visible without clipping. Exact SVG/PDF/alternative text and source quotations passed.
- Source accuracy: official ACARA workbook re-fetch HTTP 200 matched 2,049,917 bytes and pinned digest. Official current one-page QCAA Prep alignment was re-opened and all nine descriptions compared. ACARA terms repeat fetch was restricted; earlier same-session successful review is declared, not a new terms fetch.
- Tools: both original Python tools pass format/lint checks. Python 3.12.3; librsvg 2.58.0; pdftotext/pdffonts 24.02.0; pdfinfo 26.05.0; Pillow/DejaVu Sans. Default verifier performs no network calls, file writes or learner data collection.
Remaining human/local gates: educator pacing/language/adoption/timetable and curriculum decisions; actual place/view/privacy/source/task approval and responsible adult; genuine element-specific K03 evidence and permissioned actual records; individual print/screen-reader/tactile/AAC/language access review; cultural authority/ICIP for any such extension; independent classroom/parent/funder evaluation. K01/K02/K04 remain separately held. No actual care action, care benefit, classroom use, learning gain or completed K03 evidence occurred in authoring. No classroom/access/cultural/commercial certification is implied.