Evidence status: an author exercised the materials and source readings at a desk on 30 September 2026. These are simulated roles, not independent educator feedback, a child trial, family consultation, accessibility validation or a funder endorsement. No child or personal data was collected.
Educator role: I walked A/B/C into the caption task and matched marks to A, stated purpose to C and “use not recorded” to B. The first draft D originally did not name Mira, which made Day 137's misquotation edit ambiguous; D now explicitly says “Mira said”. E therefore corrects the attributed words while its final sentence keeps other uses open. I corrected the integrated-guide filenames, checked the 3+4+5+7+4+2 timing, and made ten different child products explicit. The paper museum now grows from a COPY frame and catalogue window to separate maker pockets, a caption tent, removable repair flags, a five-station trail and a reader-checked mini-exhibit. Every access route can direct this same making, while check-day folders carry no answer headings. Learner prompts were shortened and are offered one part at a time. The assessment days model only blank response mechanics before first answers. Actual pacing, teacher language, curriculum fit and local allocation still need an educator review.
Learner role: I followed the picture route, an exact read-aloud route and adult-directed shape arrangement as simulated access choices. The main strip has three bars and one hole. A/B cannot establish actual material or use; C reports why Mira drew; E does not show a model was returned. With the fresh sources, I could keep Tavi's stated shelf reminder separate from a ticket/everyone claim, and could correct Ona's badge attribution without concluding nobody wore it. That shows the author can follow the key, not that a child can. The raised-shape route was specified and walked through as placement instructions; no tactile aid was physically built or learner-tested.
Family/carer role: I checked all ten bridges for free, optional fictional play and an in-school alternative. No actual object, private memory, device, photo, translation, return slip or adult availability is required. A child can pass without assessment penalty. No family used or reviewed these bridges; fluent human translation/cultural review is needed before any translated version.
Potential funder role: I checked whether the files support the stated authored-draft inventory: ten 25-minute scripts, thirty switchable routes, twenty worked swaps, ten bridges, two fresh checks with separate key, eight original A4 pairs, exact alternatives, curriculum pins, rights ledger and a reproducible audit. They support a draft-completeness and file-reproducibility statement. They do not demonstrate learning, engagement, time saved, commercial demand, a complete Foundation year, local knowledge coverage or cultural authority.
Visual/source checks actually performed: all eight PDFs were rendered locally and inspected together. Their source marks, order, quotations, panels and footer are visible without clipping. SVG titles/descriptions, measured text bounds, A4 PDF page size, searchable embedded font text and exact visible-text alternatives are checked by the verifier. PDFs are untagged; print contrast/legibility, screen-reader reading order, tactile construction, AAC fit and classroom access remain human gates. The current ACARA workbook HTTP 200 digest matched the pinned copy; the official QCAA Prep alignment PDF opened through browsing and its nine descriptions matched. The QCAA entry's direct full-page open timed out; its current official search result was checked.
Residual critique: the source-scope task is verbally demanding for a first-year learner. Teachers should use one clause at a time and source cards kept physically separate, while preserving the same goal. Catalogue title is a deliberate ambiguity trap and must not be treated as a material label. S05 is a narrow fictional source narrative using earlier/later/after; authentic past/place work and all K01–K04 remain held. Human educator review, local adoption/consent/safety/ICIP review, individual access review and an evaluated classroom trial are pending. No “10/10” quality or outcome claim is made.
Reproduction: with Python 3, Pillow/DejaVu and rsvg-convert installed, run python print/generate_print.py --regenerate from this pack. For an intentionally refreshed receipt use python verify_pack.py --manifest-json > manifest.json; normal python verify_pack.py is read-only and uses no network. pdfinfo, pdftotext and pdffonts are required for its PDF checks. Reproduction of text/geometry is pinned by hashes; tool versions are recorded in the final verification receipt below.
Final verification receipt · 30 September 2026
- Read-only pack audit: PASS — nine exact direct workbook rows/QCAA pins; ten 25-minute scripts; 30 routes; 20 swaps; 10 bridges; two separate fresh checks; eight A4 pairs/exact alternatives/text bounds; 51 local links; 34 SHA-256 files. The manifest hashes every required pack file except itself; its separate digest can be recorded at handoff.
- Reproduction: a deliberate second regeneration produced byte-identical eight SVGs, eight PDFs and the complete alternatives file: 17/17 artifact hashes matched. Fixed PDF build metadata is a reproduction clock, not an event/source date.
- Negative audit exercise: in-memory mutations to source digest, K01 hold, lesson timing, explicit Mira misquotation, fresh-speaker isolation, learner/key separation, exact alternative text and manifest digest were each rejected at the named gate (8/8). Original text was restored in memory after each exercise. The normal audit ran before and after; all pack file bytes, sizes and modification times were unchanged by these audit exercises. No bad fixture was written into the pack.
- Source and print inspection: the current ACARA workbook matched its SHA pin; all nine local workbook rows also matched the current browsed QCAA Prep alignment. All eight rendered PDF pages were visually inspected for source marks, quotation placement, panels, reading order and unclipped margins. Automated measured bounds and complete SVG/PDF/alternative text comparison passed.
- Code checks: both original Python tools passed formatting and lint checks. Runtime: Python 3.12.3; librsvg 2.58.0; pdftotext/pdffonts 24.02.0; pdfinfo 26.05.0; DejaVu Sans fonts. No network, file rewrite or child data is part of a normal verifier run.
The block is an authored, reproducible teaching draft awaiting human educator, local curriculum/consent/safety/ICIP, individual access and classroom trial review. The audit is not one of those reviews and establishes no child learning outcome.