# Four-role desk run-through · W05–06 improvement

**Actual review type/date:** internal author-agent desk review on **2026-09-30**, with actual reading of scripts/routes/source/key, direct and indexed primary-source checks, complete PDF rendering and inspection, independent signal reconstruction, temporary negative probes and temporary byte reproduction. These roles are analytical viewpoints, **not actual teacher/parent/learner/funder participants**. No classroom pilot, human hearing, device/volume clearance, child response or independent accessibility test occurred. The separate publishing-agent review below is complete.

## Teacher · start Day 21 with one phrase

The README now opens with a tiny picture-show opening, puppet ending, music alone or own child purpose. One familiar locally reviewed quiet source or four drawn cells is enough. Cue four comfortable beats; a willing child tries rhythm at 1, 3 and 4 once before looking at A. They can instead direct actual adult sound or plan silently. The teacher asks where the rhythm waited, then offers keep/change/stop. Counting follows the trial only if useful. Pulse alone comes first if wanted; no compulsory simultaneous duet or fixed performance speed. The ten full scripts retain25 minutes while allowing one phrase/another sitting.

**Concrete desk decision:** add immediately usable familiar source/stop instructions; distinguish actual child sound, actual directed-adult sound and silent direction in the header/private evidence note. For Day 24/27/28, a tap cannot sustain LONG: offer a suitable reviewed wanted hold or record a separate short/rest trial and a held plan. An optional comfortable hum or already available familiar sustain control can be tried; neither is required. Printed held bands do not certify acoustic duration. No new formal assessment or recording duty.

## Learner · choose a Day 22 ending

The child may try quiet-middle or quiet-end, only one phrase if enough, for their own ending or music alone. “Keep it, change it, or stop?” asks an actual artistic choice, not a formal score audit. Broad fixed cells, drawing, pointing, AAC and neutral adult placement remain switchable options; they are not assigned by diagnosis or learning style. Waiting belongs to the part and never requires bodily stillness or suppresses communication. A private plan, actual adult rendering or child quiet phrase can each be complete for its stated evidence.

**Concrete desk decision:** shorten ordinary prompts; move dated maker/help/aural/pulse/rhythm/audience notes into teacher sections. Open alternatives remain valid music while the particular planning target may be not yet observed. Fresh K/L brief/data/key remain unchanged; unfinished at the time limit is not yet observed and another short sitting is available, with no completion-speed score.

## Parent · one free invitation

A caregiver can choose one home bridge or none: a tiny drawn opening/ending, four broad cells, or a familiar locally reviewed quiet phrase if wanted. One minute, music alone and stopping are allowed. No instrument/device/purchase, recording, return sheet, adult report or family practice disclosure. The same invitation is available at school. Candidate media are excluded until human listening/device/volume clearance; home participation does not depend on those files.

**Concrete desk decision:** turn home rows into short invitations rather than repeated paper counting questions, keep all 20 free/optional, and label Day 25/30 offers **After Check** with unrelated practice rows. No home evidence is inferred from an invitation or returned drawing.

## Assessor / potential funder · inspect what is supported

The canonical workbook's actual code/text cells and imported Foundation Music records match all four crosswalk descriptions. The current QCAA official indexed alignment confirms aspect mapping; direct 403 is recorded honestly. Little Window Rhythm supplies an accurately named SubjectNest computer-created model/source dated 2026-09-30, with documented author purpose and original sound recipes. Actual encounter plus own observed child response is conditional/narrow FE01 evidence, not broad people/community coverage. Varied appropriate additional works remain necessary. Source-card purpose is not an achieved effect. Actual audience share needs a willing real recipient and medium; imagined/paper puppet rehearsal does not qualify.

**Concrete desk decision:** preserve the fresh K/L key and print bytes; save a baseline and complete transcripts; verify actual printed event positions, held bands, signal components and all declared silence. Temporary probes reject wrong same-count/fresh-leaking rhythms, timing shift, missing part, incomplete transcript, false human attribution, silent direction labelled adult performance, altered fresh key and two-page aid. Byte hashes/reproduction establish integrity, not classroom effectiveness. No whole-year/whole-curriculum certification from200 authored starts or pack counts.

## Actual rendering and held gates

All **eight PDFs** were rendered with Poppler at 100 dpi and visually inspected in three full-page groups: revised A/F; unchanged B/C/D/E; fresh K/L. All are one-page searchable A4 with embedded DejaVu, broad numbered cells and no observed cropping, clipping or word overlap. The revised A names its computer-created creator/type/date and removes child count blanks; F shows own purpose, actual trial category and optional share with private teacher notes. Exact alternatives match every printed word and the full rhythm/hold arrangement.

Numerical checks do not establish real hearing, tactile fit, classroom timing, child comfort, human educator judgment or accessibility certification. Local staff still review live materials/access/stop, and must listen on intended equipment/set comfortable volume before any candidate playback. Actual source encounter, child making/responding and audience sharing remain unobserved until they happen. Broader repertoire/cultural authority, classroom feedback and independent accessibility review remain open.

## Independent publishing review · 30 September 2026

A second agent read the revised daily instructions, material/source route, participation choices, worked examples, held checks, exact crosswalk, rights and review records from the frozen copy. The ordinary Day 21 trial uses rhythm sounds at 1/3/4 and a rest at 2; pulse is a separate continuing part. Day 24's unavailable-hold route makes a separate child-selected short/rest phrase and keeps the held target as a plan. Day 29 offers an actual score encounter and optional real recipient without inventing child performance or an audience response. Day 30's new held score remains distinct from ordinary practice. These are desk interpretations, not learner observations.

The publishing agent ran the default audit, all ten temporary negative probes and 24-file byte reproduction successfully. Both changed stored A/F PDFs were independently rendered and viewed in full: the source caption, two aligned rows, broad own-work cells and optional purpose/actual-try/share fields are readable and within the pages. Unchanged fresh assets and key remain independently byte-pinned. Current official indexed QCAA alignment and ACARA Arts overview were separately read; no new direct QCAA download or physical printing is claimed. The unmeasured “ready in a minute” heading was changed to “Quick setup”. No remaining blocker was found for publishing this optional teaching draft. All human audio, classroom, individual access and broader repertoire gates above remain open.