# Teacher, child, assessor and funder desk run-through · Weeks 23–24

**Teacher desk run:** Day 113 fits 25 minutes (3 source, 4 before, 5 during, 7 after, 4 check, 2 exit). The adult pretests a tiny drop inside a level catch tray well before the session; if there is no approval, clear material identity, accessible observation or safe cleanup, use D as FICTIONAL MODEL and write REAL OBSERVATION NOT AVAILABLE. Day 114 may collect a second actual pair only if safe and feasible; never rush, complete missing real entries from D, or hide run-off. A real accidental spill interrupts the public exercise and follows adult school procedure.

**Child desk run:** D's pretend paper entries are DARK PATCH twice; the plastic entries show a ROUNDED and a FLAT VISIBLE DROP; a third plastic entry ran to the edge and is INTERRUPTED. I can say that about a *fictional record*, and I cannot see where the water went. I may speak, sign, use AAC, point, draw, hear an attributed ADULT REPORT or use broad secured tactile paper labels. I need not touch water or tell a personal event/feeling. If no real test occurred, I can say we have not observed it yet.

**Assessor desk run:** Held Z's DISPLAY PANEL 27 and changed quiet-count waits, plus held AA's READING DISPLAY MODEL 84 and sleeve-covered row, do not appear in instructional D/E, learner routes, worked swaps or print aids before first response. First wording, access and neutral prompts are recorded before the separate key. Paper-only cannot credit AC9SFI02 actual investigation, a real material property, water destination or universal rule. No prior prediction is invented. Two paper checks do not certify year/term mastery or child growth.

**Funder/public desk run:** Seven exact pinned ACARA rows, QCAA Prep entry, original CC BY texts/vectors, clear wet-activity gate, no-cost family route, rights/privacy limits and SHA manifest are inspectable. Educator, accessibility, local syllabus, material/safety, privacy and classroom review remain due. No field pilot, measured effectiveness, local cultural authority, translation or endorsement is claimed.

**Mechanical/visual receipt:** [Verifier](verify_pack.py) checks ten six-step 25-minute scripts, 30 same-target routes, 20 worked swaps, ten family prompts, two fresh checks/keys, seven official rows/workbook hash, local links, CommonMark list spacing, six A4 SVG/PDF pairs and file hashes. Markdown-it rendered learner cards as ten lists with thirty items and optional practice as ten lists with twenty items; a headless browser preview of the learner page was visually checked for real bullets and clear headings. All six PDF pages were visually inspected. This is a desk review, not a classroom trial.