# Internal teacher, learner and funder run-through · Year 7 maths

**Status:** internal simulation on 29 September 2026. No class, child, external funder, approved local assessor or accessibility user was observed. This documents desk reasoning and issues, not evidence of uptake, learning gain or formal curriculum coverage.

| Role/task simulated | What worked | Critique and response |
| --- | --- | --- |
| Teacher prepares Week 3 | The six-step script totals25 minutes, gives three route choices/day, and extends the opening fortnight into multiplying fractions and decimal grouping division. | An operation answer without whole/unit can be meaningless. The mat and exit prompts require whole, unit, estimate/inverse. |
| Learner tries Day 11 and13 | One same-size whole supports `1/2+3/8=7/8`; a 3-by-4 overlap supports `3/4×2/3=6/12=1/2`. | “Multiply the denominators” alone could mask misunderstanding; model asks why the result is smaller than both proper factors. |
| Learner tries Day 14 | `2.4÷0.6=4` counts intervals, while five endpoint markers may be drawn. | Unit confusion could produce “four units”. Prompt/key state **four intervals**, with multiplication check. |
| Teacher administers Day15 | New numbers and contexts are not used in prior learner routes; three items separate equivalence, grouping and percent-of-whole. | Ten minutes may be tight for some learners. Neutral read-aloud/blank tactile route or a later valid completion is allowed and recorded. |
| Teacher prepares Week 4 | The factor tree has space for up to eight leaf boxes and a multiply-back line. Different 72 trees end with the same `2³×3²`. | Initial last-level boxes visually touched. The drawing was revised to leave visible gaps, then regenerated and visually inspected. |
| Learner uses Day19 square link | 12-by-12 blank grid can support144; the positive root is checked with12×12. Prime factor counts also verify144=`2⁴×3²`. | Counting144 tiny cells is not required; spoken multiplication or large tactile cards are an equal mathematical route. |
| Learner takes Day20 check | New108,196,75 require prime leaves, exponent count, square/root inverse and error checking; separate key has exact products. | The prior Day17 extension initially mentioned108, which would leak a check value. It was changed to36 versus72 before handoff. Fresh-number scan is in the verifier. |
| Potential funder inspects reuse | Original vector aids, text/tactile routes, 30 practice routes, ten extras, separate checks/keys, exact ACARA row crosswalk and file hashes are inspectable. | No classroom effect, uptake, approval, independent accessibility result or financial return is claimed. The README/ledger state that boundary. |

## Desk arithmetic and print audit

Exact operations were independently recalculated in the verifier with `fractions.Fraction` and decimal arithmetic; factor powers multiplied back and positive roots were squared. The three one-page A4 PDFs were rendered and visually inspected for clipping and spacing; `pdftotext`, `pdffonts` and `pdfinfo` verify selectable text, embedded DejaVu subsets and A4 dimensions. The text alternatives give full prompts and blank structure. Automated checks and one-person visual inspection do not replace user testing.

The [read-only verifier](verify_pack.py) also checks ten lesson headings/timings, 30 routes, ten extras, six fresh items, held-out learner-detail leakage, local links, exact pinned descriptions and [hash manifest](MANIFEST.sha256). Next iteration should use teacher/learner evidence to revise pacing, then map each target Australian jurisdiction before local release.