Status: author/desk simulation on 2026-09-29. No child, teacher, family, community custodian, external assessor, physical printer or real audience was involved. The scenarios below test internal coherence, not classroom efficacy.
| Lens | Run-through | Finding / correction |
|---|---|---|
| Educator, Day 21 | Trace START→A and START→B using A4 map. | A fork is explicitly separate from a three-frame line; both outcomes need not be seen by one reader. |
| Educator, Day 23 | Select a sign with prebuilt covers, then reset. | PDF is static; physical broad covers or removable cards are required for an operable board. Route documents that boundary. |
| Educator, Day 26 | Can a learner with limited hand movement still author the work? | Adult can attach child-selected cards/hinges as a neutral fabricator; record child choices and adult construction separately. |
| Educator, Days 25/30 | Do public checks accidentally rehearse solutions in A–F? | New I triangle branch and J sign-matching source stay out of practice; same-day swaps are After Check. |
| Learner, low print reading | Can A/B be understood without alphabet or colour? | Choice signs can be broad shapes/textures, read aloud or traced by AAC; no independent reading needed. |
| Learner, sensory/privacy | Must a child reveal, touch, film or perform publicly? | No. Child may direct adult operation; sound/device absent; Day 29 private rehearsal/later/pass remain separate from sharing evidence. |
| Funder / assessor | Does “made a media work” overstate production? | An assembled child-directed paper choice object is small paper media making. A drawn fork alone is planning. Neither demonstrates camera/audio/digital work or audience understanding. |
Self-critique: The new interactive paper format adds an authentic maker choice but is materially simple. Some learners need larger boards, secure broad parts or a partner for tactile access. The 25-minute assembly may run long in a particular class, so a teacher should prepare neutral physical components beforehand and observe actual timing. The desk review cannot prove a child will grasp branch logic or that a reader will operate the paper work as intended. Trial feedback should check ambiguity of signs, resetting after a branch, whether the first check is genuinely uncoached, and whether a child’s intention is preserved when an adult builds the hinge. The real-work slot remains unresolved until the school verifies permission and cultural authority; fictional shapes are not a substitute.
Technical QA planned/recorded at freeze: python generate_assets.py --write to generate original aids, python verify_pack.py read-only for structure/links/hash and PDF checks, ruff check generate_assets.py verify_pack.py, SVG text versus exact alternative comparison, asset regeneration byte comparison, and rendered A4 page inspection. These checks cannot establish material safety in a specific classroom, copyright permission for external works, learner outcomes or actual audience behaviour.
Observed desk/technical result: The eight rendered A4 pages have distinct branch and choice-sign shapes, contained labels, visible hidden/revealed states, and unsolved I/J source diagrams. No clipped text or solved answer panel was seen in the contact-sheet inspection. All eight SVG text streams matched their generated verbatim alternatives; repeat generation produced identical SVG/PDF bytes. The read-only verifier and Ruff passed after the source, lesson and assembly-instruction review. A physical board has not been printed or tried by a learner, so paper-handling usability and timing remain unverified.