Teacher route. I walked Days 31–35 as a new teacher. Source A is clearly an original practice tag; the public materials deliberately leave real sign text, placement and action blank. To teach AC9HPFP06 locally, I need the school child-safety lead to verify two current accessible real signs and the correct action on a dated school copy. If none is available, the lessons still teach source distinction and clarification, but the record says REAL COMMUNITY SIGN NOT OBSERVED. Case J tests that distinction with a new tag, not local sign mastery. Any real concern leaves the lesson and follows school policy. This is a desk rehearsal, not a school visit.
Child and access route. I walked a no-speaking, no-touch, no-sound, no-physical-performance route through Days 34 and 38: an adult reads the full source, the learner directs broad cards or AAC, and the same clarification/stop reasoning is available privately. Week 8's physical component remains open for an individually suitable, locally checked action; paper direction is not misrecorded as child movement. A child can say no, pause or pass without giving a reason. An educator must still test actual AAC, tactile choice, real sign access, stop cue and physical setting with the learner.
Freshness and evidence route. Case J's tag and Case K's strip are distinct from teaching and worked swaps. I answered each by its exact lines, then checked the separate key. The script, check and key all use the same Day35/Case J and Day40/Case K pairing; the verifier fails a mismatch. Case K's paper REST answer does not certify HPE movement, and adult demonstration stays adult evidence. A fresh public case is formative, not a secure exam.
Funder and rights route. A reviewer can trace exact ACARA rows, the QCAA entry, dates, licences, original vector hashes and explicit evidence gates. The claim is a coherent draft resource with no copied sign, child image, cultural movement or paid worksheet. There is no measured safety improvement, physical gain, uptake, certified accessibility, classroom pilot or authority endorsement. Human educator, safeguarding, sign-rights, local access and trial review remain needed before stronger claims.
Mechanical QA on freeze. The deterministic verifier checks ten 25-minute six-step scripts, thirty A/B/C routes, twenty worked swaps, ten optional bridges, exact held-case IDs across files, fresh phrases, official workbook rows, six A4 SVG/PDF pairs with selectable exact words, links, Markdown list spacing and SHA hashes. Ruff, rendered learner lists and visual PDF inspection are separate. These checks cannot establish a real sign's meaning or that any child moved.