# V7 — INDEPENDENT RE-VERIFICATION, BASELINE v2.0.0

I built none of this and I took no docstring, comment or checker PASS line on trust. I read the 32 requirements out of `project_data.REQ` and then opened the artefacts. I extracted all 178 top-level paragraphs and all 45 tables of the delivered study with python-docx, cell by cell; I opened the delivered workbook twice with openpyxl, once for formulas and once for cached values, and dumped all fourteen sheets; I opened the Decision Pack with python-pptx and read every run, every table and every picture shape on all twelve slides. I re-derived the whole cost-benefit case from `model_params` and `model_agedcare` in my own arithmetic without calling the builder's `build()`, and got $420,113.58 of benefit, $275,726.60 of cost, a net of $144,386.98, an ROI of 52.366% and break-even at month 17 / month 10 — then compared that against `business_plan.json`, against the A3 page, against `trade_study.py`, against ASM-028 and ASM-029, and against the study. I diffed the delivered `business_plan.json` field by field against a fresh in-memory rebuild. I printed `BUSINESS_PLAN_A3.html` in headless Chromium five ways, rasterised two of the resulting PDFs at 100 dpi and looked at them. I ran my own sub-10pt scan across every Office file in `05_Outputs`, including inherited paragraph-style sizes and raw `sz=` attributes that the project's own check cannot see. I ran my own jurisdiction scan, my own secrets and PII scan over 102 files, and my own attribution measurement at two different strictnesses. I viewed `D11_budget_constraint.png` and `D12_two_entities.png` as images. I read `SYSTEM_BREAKDOWN.md`, `SOURCE_MANIFEST.json`, `Dashboard.html`, `README.md`, `09_Help_Hub/index.html`, `00_Handover/`, all of `03_Registers/*.csv`, both T3 records, and the source of C05, C10, C12, C22, C27, C29, C31, C33, C34, C40 and C41. I built a throwaway copy of the whole tree in scratch and attacked the new C41 baseline mechanism in it. I did not run `build_all.py`. I did not run `run_checks.py` inside the project — I read the log the builder's own run left at 07:07:57 and ran the suite only inside my sandbox copy — so `checker_run_log.txt` is untouched and I left no `__pycache__` behind.

Two things dominate this pass. First, most of the eleven fixes are real, and two of them (DEF-027 and DEF-030) are better than the defect required. Second, **the source of truth is now ahead of every artefact that projects from it.** `01_System/project_data.py` was last written at 07:07:36; every register CSV was written at 07:05, every deliverable, the dashboard, the README, the Help Hub and the A3 page at 07:06. So DEF-025 to DEF-034 exist in the source and in none of the 24-row defect registers; the study's Appendix B2 still publishes "24 defects, all closed"; and the dashboard's headline assurance tile still reads **26 / 32 requirements INDEPENDENTLY verified**, sourced to a 20 August pass, while the source of truth now derives all 32 as Open. That is the exact condition DEF-035 was raised to make impossible, still visible on the surface a stranger opens first. C40 catches it and fails; nothing else does.

## Did the V6 fixes hold?

| Defect | Fix claimed | Verdict | What you actually checked |
|---|---|---|---|
| DEF-025 | The A3 page prints to one A3 landscape sheet; C34 now counts pages in a real PDF | **NO** | It prints to one page, but only because `@media print{.sheet{height:296mm;overflow:hidden}}` throws away whatever does not fit. I ran C34's exact call (`pg.pdf(prefer_css_page_size=True, print_background=True)`, no format) in a fresh browser three times: 1 page, 420×297 mm — and I rasterised it. **The page ends part-way through the "8 · AREAS OF CONCERN" heading. Section 8's table, all of section 9, the entire value-validation verdict block and the footer are gone.** Text extraction: 10,087 chars against 16,726 for the same file at an explicit A3 size. Then I quadrupled the card content in a scratch copy and ran C34's call again: still exactly 1 page, still PASS. **C34 cannot fail on content.** At `format='A3', landscape=True` the page does render in full, three columns, legibly, with margin — I looked at that render too — so the layout is capable; what is broken is that the file's own `@page` rule plus `@media(max-width:900px)` collapses it to one column at print layout width and `overflow:hidden` hides the consequence. |
| DEF-026 | The three-year net is computed, not typed, and consistent everywhere | **NO** | `$145,621` is gone. It has been replaced by a **new** inconsistency. `trade_study.py` now calls `build_business_plan_json.build()` at run time and emits **$144,386.98**; `business_plan.json` and the A3 page carry $144,386.98 twice each. But ASM-029 — in `project_data.py` line 1579, `03_Registers/ASM.csv` and `00_Handover/ASM.csv` — says the computed net is **$144,421.10**, and `trade_study.py`'s own docstring says $144,421.10. ASM-028 states the $90k sensitivity against "$420,147.70 of benefit … a net of $54,421.10"; the delivered benefit is $420,113.58 and the net at $90k is $54,386.98. The DEF-033 rounding fix moved the number and nothing downstream followed. Two artefacts still state different values for the same named quantity; the gap shrank from $1,199.90 to $34.12. |
| DEF-027 | Weights and results are separate files; the weights file states what it can prove | **YES** | Two files now. `T3_trade_study_weights.md` retracts its own prior attestation in terms: "that sentence is an attestation, not evidence … Independent verification V6 was right to reject it", then says plainly what was actually done and concludes "The ordering is attested, not proven", offering the live `TradeStudy3` sheet as the real mitigation. That is the correct answer to an unprovable claim and it is the best piece of writing in the project. I checked the results file's sensitivity table against `trade_study.sensitivity()` — 0.31→C(3.61), 0.28→C(3.63), 0.23→D(3.52), 0.26→D(3.64), contribution never flips — exact. One flaw: the weights file names criterion 5 "**Dependence** on the owner's own hours" while `trade_study.py`, the workbook and the study all name and score it "**Independence** from the owner's own hours". The GOV-B6.3 record states the fixed criterion with inverted polarity against how it was scored. |
| DEF-028 | C29 opens all three Office artefacts; the Standard's exclusion is derived, not hard-coded | **PARTLY** | C29 does open all three, and the exclusion is derived from the CI register's owner column rather than a filename — correct and I confirmed the mechanism. The artefact side is genuinely fixed: I re-scanned independently and found **zero** sub-10pt runs in the .docx (including inherited paragraph styles — three styles are defined below 10pt, `Body Text 3` at 8pt and `Caption` at 9pt, but no paragraph uses them; effective sizes are 10.5/13/17), **minimum 10.0pt** across 194 runs in the .pptx against V6's eight 9pt runs, and **zero** cells below 10pt in 1,229 workbook cells against V6's 145. The exclusion of `AI_Project_Governance_Standard_v3.7.docx` is justified — it is Zaid's input document, CI-001, filed under GOV-F5.7 — and it is load-bearing: I measured 1,450 sub-10pt runs in it. What did **not** get fixed is the six-column half. Scenarios is still 14 columns, Costs 7, TradeStudy3 7. That now passes on DEC-008, whose own record reads "**NOT RATIFIED BY ZAID**". A requirement satisfied by an unratified reinterpretation is not satisfied. |
| DEF-029 | Help Hub sentences generated; dashboard shows combined capital; no false statements left | **PARTLY** | Both sentences V6 named are fixed and now true: "Every artefact this project generates carries the currency date 2026-09-07 and baseline v2.0.0", and "gathered in 2 waves: 53 on 2026-08-20, 21 on 2026-09-07" (I counted the SRC register: exactly 53 and 21). HLP-03 now reads "24 raised, 24 closed". HLP-05's fifteen-minute route now crosses both parts (sections 1, 15.3, 15.2, 16, 5.2, 18, 19) and says "Seven findings". The dashboard has a $44,483 COMBINED capital tile, a $11,492 working-capital tile, and the $15,875 NDIS-only figure demoted with "Read it beside the combined figure above, not instead of it", plus $1,002/$435 aged care tiles. **But the same two surfaces still carry at least eight false statements of exactly the class that was fixed** — see New defects D-V7-4. The most serious: the dashboard's headline "26 / 32 Requirements INDEPENDENTLY verified" and the study/README/dashboard ACT-013 claim that at $90,000 "the three-year return falls from 53 per cent to roughly 10 per cent", which is 52.4% to 14.9%. |
| DEF-030 | The three false workbook-README statements are gone; the currency sentence is true | **YES** | Title is now "…Financial Model  v2.0.0". The currency sentence is "74 sources across 2 research waves: 53 verified 2026-08-20, 21 verified 2026-09-07 … the FIRST source to expire does so on 2027-08-20" — I checked it against the SRC register's accessed-date histogram: exactly 53 and 21, first expiry 2027-08-20. True. The blanket "nothing is typed in" is now "Nothing … is a typed-in result EXCEPT the two comparison sheets named below", naming Structure B5:C6 and Combined B5:C9. I scanned every non-formula numeric cell on every sheet outside Inputs/Drivers/AC_Inputs/README: the only typed **results** are exactly those declared cells. Everything else typed is an input (fees, budget levels, hour ramps, utilisation steps, trade-study weights and scores). The claim is now accurate. 455 formulas, all 455 carrying cached values. |
| DEF-031 | SRC-074 exists and contains the figures | **PARTLY** | SRC-074 exists in `project_data.py`, `03_Registers/SRC.csv`, `00_Handover/SRC.csv` and study Table 20, and its Claim field carries "$63.85 per client per day against direct cost $47.75 … a positive margin of $16.10 per client per day". **But nothing cites it.** Every place the figure is actually used still cites SRC-068, the row that does not contain it: study finding 5 "[SRC-068, SRC-069, RSK-011, V4]", study Table 29/15.1a "[SRC-068, SRC-069, RSK-011, V4]", RSK-011 "about $16.10 per client per day **[SRC-068]**", deck slide 9 footer "SRC-068 sector results". SRC-074 appears in exactly three files, all of them register projections, and in no citation anywhere. The row was created and never wired up. |
| DEF-032 | The classification table starts at $10,731 with 1.80 hr/week | **YES** | Study Table 15.2 row 1: `$10,731 a year [SRC-064] \| 1.80 hr \| $894.25 \| $302.14`. Workbook `AC_ClientEconomics!A22 = 10731` with B22/C22/D22 live formulas resolving to 1.80126899577001 / 894.25 / 302.14339725942. Deck slide 9 table row 1: `$10,731 a year \| 1.80 hr \| $894 \| $302`. Diagram D11's first bar is labelled `$10,731 / L1` at 1.8 hr / $302 pcm, and its source note records the correction by name. Propagated to all four surfaces. |
| DEF-033 | The placeholder is gone; the delivered JSON reconciles | **YES** | I read `build_business_plan_json.py` end to end. The `and NR and 0 or 0` placeholder and the 250-lines-later repair are both gone, replaced by a comment saying why. `ndis_contrib_hr_paid` is now recomputed from `price − loaded_wage(l2) − admin × loaded_wage(l3)` = 6.69692875 rather than the rounded 6.70. I diffed the delivered `business_plan.json` field by field against a fresh `build()` — **zero differences**. And I re-derived the whole CBA myself without touching the builder: 11,110 NDIS hours × 6.69692875 + 345 aged care client-months × 1,002.06 = **$420,113.58**; costs 21,948.60 + 66,258.00 + 7,520.00 + 180,000.00 = **$275,726.60**; net **$144,386.98**; ROI **52.366%**; break-even month 17 with opportunity cost, month 10 on cash. Every one matches the delivered JSON to the cent. |
| DEF-034 | Tables 4.2, 7.0, 10.2 and the WorkCover/LSL/WWCC roadmap rows exist; Part II carries diagrams | **YES** | Table 4.2 exists: 12 rows × 6 columns (Obligation / Authority / Cost / Frequency / Lead time / Source), **zero empty cells**, every row carrying a SRC — I tested cell by cell, as I did for Table 4.1 (17 rows, zero empties, frequency column present). Table 7.0 exists and answers time-to-first-revenue **and** cost-to-be-ready for all four models, with the support coordination row saying explicitly that its $0 "is the absence of a cost, not a cost of zero" and pricing what happens if the pause lifts. Table 10.2 exists: nine sequenced aged care rows. Table 10.1 row 6 is a new "Weeks 5-7 / **BEFORE ANY PAY RUN" row registering WorkCover and portable LSL, and row 5 now carries the WWCC. Part II carries D11 (section 15) and D12 (section 17); 12 diagrams on disk, 12 `word/media/*.png` embedded. I viewed both PNGs at full size: D12 is clean, nothing clipped, nothing overlapping; D11 is legible but its orange "the modelled average, 5.04 hr [ASM-020]" label is drawn on top of the dashed line it annotates rather than above it. |
| DEF-035 | Every verdict carries a baseline; C41 fails on a stale one | **NO — I defeated it in five minutes** | The shape is right: `IV_VERDICT_RAW` is `{rid: (result, baseline)}`, `IV_VERDICT` downgrades anything whose baseline ≠ `BASELINE_VERSION` to STALE, `REQ` status derives from that, and today it correctly holds all 32 at Open with 26 marked stale. But C41 tests only `os.path.exists()` on the file named in `IV_SOURCES["v2.0.0"]` — **it never opens it**. In a sandbox copy of the whole tree I wrote a file at `02_Work/scratch/V7_baseline_v2_reverification.md` whose entire content is "EVERY REQUIREMENT FAILS. 0 of 32 requirements PASS.", set `IV_V7 = {r[0]: "PASS" for r in REQ}`, and ran the suite. **C31: PASS, "32 of 32 requirements independently verified". C41: PASS, "32 of 32 requirements carry a verdict taken against the CURRENT baseline v2.0.0". Total: 1 of 41 checks failed, and that one was C40 complaining about a byte count.** Then I appended 4 KB of nulls to the delivered study and re-ran: C41 still PASS. Second hole: `BASELINE_VERSION` is a hand-typed string, not a content identity — every artefact was rebuilt at 07:06 under an unchanged "v2.0.0" and the source of truth edited again at 07:07 under the same string, so a "v2.0.0 verdict" does not name an artefact set. Third: C41's second condition (`result == "PASS" and baseline != cur and status == "Verified"`) is unreachable, because `status` is only "Verified" when the baseline already matches. The check has one live test, not two. |

## Verdict table

| Requirement | Verdict | Evidence you personally opened | If FAIL, exactly what is wrong |
|---|---|---|---|
| REQ-SYS-01 | PASS | Study Table 5.1 gives core supports $9,110–$12,786 (base $10,791), support coordination $3,673 and SIL/SDA $8,600–$16,900, each row citing SRC-025 to SRC-041 / DEC-005; Table 18.1 gives the aged care $11,157.80 and the combined $21,948.60. I traced the NDIS lines to workbook `Costs` B6:D25 and the aged care one-off to `model_agedcare.ONE_OFF_COSTS`. SIL/SDA carries DEC-005 and is named in the study. | — |
| REQ-SYS-02 | PASS | Study Table 7.0 (section 7.1) gives a month-count for **all four** candidate models: core supports "About 4 months" [ASM-015], support coordination "About 2 months … the fastest line in the study" [SRC-018], SIL/SDA "9 to 12 months at the earliest" [SRC-021], aged care "About 4 months, on a registration whose lead time is not published — which is why ACT-011 exists" [SRC-065]. No model states a range with no source. | — |
| REQ-SYS-03 | PASS | DEC-004 records weights at 2026-08-20T09:40 and scores at 11:05, in separate register fields. Study Table 6.1 shows three alternatives; I recomputed A1 2.85, A2 4.05, A3 2.30 by hand and they are correct. Sensitivity names the flipping weight (addressable market above ~55%); rationale present. | — |
| REQ-SYS-04 | PASS | Study Table 7.1: core supports 3.30, support coordination 3.70, SIL/SDA 2.20 against six identical weighted criteria; I recomputed all three. Table 7.1a states the flip (regulatory stability 25%→40% gives 3.45 v 3.40), which I verified arithmetically. | — |
| REQ-SYS-05 | PASS | Workbook `UnitEconomics!B11 = SUM(B5:B10)` → 21.30424, where B5 = `Inputs!$B$5` (SRC-002, 73.58) and B6:B9 are `-Inputs!$B$8` times `Inputs!$B$10/$B$11/$B$12` (SRC-009, SRC-011, SRC-012, SRC-013). Not one constant in the chain. | — |
| REQ-SYS-06 | PASS | Study Table 4.1 (NDIS entity, 17 rows) and Table 4.2 (aged care entity, 12 rows), both six columns — Obligation / Authority / Cost / **Frequency** / Lead time / Source. I tested every cell of both tables programmatically: **zero empty cells**, and every data row's last column carries a SRC-###. Aged care worker screening (SRC-057) and the ACQSC registration (SRC-065) both now appear, the latter with its lead time recorded honestly as "NOT PUBLISHED … ACT-011 closes it". | — |
| REQ-SYS-07 | PASS | Workbook `Scenarios` rows 37-45: the downside scenario carries a formula-driven month-by-month cumulative line from `-11,638.13` to `-20,958.80` and a single runway figure `N45 = MIN(B45:M45)` → `-20,958.80`, which reconciles to 10,790.80 + 847.33×12 as I computed independently. Study Table 8.2 row 4 states -$20,959. | — |
| REQ-SYS-08 | PASS | Study Table 9.1 ("WHAT WOULD CHANGE THIS ANSWER") lists five items, each naming a register ID — ASM-008, ASM-002, SRC-002, ASM-015, SRC-049 — and three name the checkable party ("a real provider", "you", "government publishes"). Part II adds a second such box in section 19. | — |
| REQ-SYS-09 | PASS | Study Table 10.1 (13 rows) and Table 10.2 (9 rows), both under a heading that says "This roadmap is CONDITIONAL". I walked every obligation in Tables 4.1 and 4.2 against them: WorkCover and portable LSL are now sequenced at "Weeks 5-7 / **BEFORE ANY PAY RUN" in both entity roadmaps, the WWCC at Weeks 5-8, and the two that are not sequenced (voluntary disability worker registration, the annual company review) are named in the caption with the reason. | — |
| REQ-SYS-10 | PASS | `BreakEven!B6 = Costs!$C$28/UnitEconomics!$B$16` → 39.77; `B7` → 126.53; `SupportCoordination!B13 = Costs!$C$28/Inputs!$B$7` → 8.46; aged care 1.0 client from `model_agedcare.breakeven_clients()`. Every figure resolves through input cells. SIL/SDA excluded by DEC-005 and named in Table 5.1. | — |
| REQ-SYS-11 | PASS | `BreakEven!B12` is a live formula over `Drivers!$B$14`, the loaded-wage chain and the ASM-007 payment lag → 7,318.61, with B13 giving 15,682.73 at 30 days. Aged care $4,173.20 from `model_agedcare.working_capital()` at a seven-day lag (ASM-030). No constant anywhere. | — |
| REQ-SYS-12 | PASS | Study Table 7.0, fourth column, answers the July-2027 question for **all four** models and pairs each with a cost to be ready: core supports "Mandatory registration arrives. Trading unregistered stops being lawful" at $3,500–$6,000 + $649–$1,825; support coordination "The exposure here is redesign, not registration" at "$0 audit and $0 manual TODAY, because mandatory registration for support coordination is currently PAUSED — and that is the absence of a cost, not a cost of zero", with the cost to be ready if the pause lifts priced explicitly; SIL/SDA "Already in force. There is no unregistered pathway to close" at $8,600–$16,900; aged care "Not affected by the NDIS change" at $600–$3,000 + $4,997. No model is silent, and the V6 gap (support coordination carrying no cost to be ready) is closed by naming the absence rather than by inventing a figure. | — |
| REQ-SYS-13 | PASS | 12 PNGs in `02_Work/diagrams/` and 12 `word/media/*.png` in the delivered package. Part II now carries D11 at section 15 (the fixed-budget mechanism) and D12 at section 17 (the two-entity dependency structure) — the two process narratives V6 found in prose and tables alone. I opened both images: D12 renders cleanly, nothing clipped or overlapping; D11 is legible, with one label drawn over its own reference line. | — |
| REQ-SYS-14 | PASS | I read the HLP-10 table in `09_Help_Hub/index.html` row by row. All nine GOV-G3.4 symptoms are present as real symptom/cause/check/fix rows: dead link, chart empty or wrong, dashboard blank or will not load, register will not open, two documents disagree, #NAME?, wrong version, phone view, conflict copy. Count = nine, and none of them tells the reader to ask anyone first. | — |
| REQ-SYS-15 | PASS | `Dashboard.html` carries "GENERATED FILE … Rebuilt from `01_System/project_data.py` by `01_System/build_web.py`"; it is the only dashboard in the tree; all twelve tile values project from register or model values. Nothing is typed onto the page. | — (but see D-V7-1: it is a faithful projection of a source that has since moved, and its headline tile is now false) |
| REQ-SYS-16 | PASS | I extracted the twelve `data-metric` attributes from the dashboard myself and matched them against `03_Registers/METRIC_LINEAGE.csv`: 18 lineage rows, **zero** tiles without a row, and **zero** empty cells across all 18 rows × 6 fields. Metrics-without-lineage = 0. | — |
| REQ-SYS-17 | PASS | The builder's run log shows C22 "15 pack files re-derive from their sources" and C23 "C22 detected a deliberate one-line tamper in HANDOVER.txt and the file was restored". I reproduced both in my sandbox copy. The negative test demonstrably fails on demand. | — (but see D-V7-6: C22 compares SHA-256 against a manifest written at the same moment; it does not re-derive anything, and the pack it certifies is stale) |
| REQ-SYS-18 | PASS | I parsed `03_Registers/SRC.csv` myself: 51 of 74 rows sit below High confidence (42 Medium, 9 Low) and the 'Why below High' field is non-empty on all 51. I tested for emptiness, not for the checker's word. | — |
| REQ-CON-01 | PASS | C05 reports 0.966 over 292 material figures with 10 unmatched; I opened all ten and every one is internal arithmetic (123 = 45+78, 1,849 = 649+1,200, 5,325 = 1,825+3,500, 406.39 = 87.64+318.75, 6,797 and 9,997 = the aged care policy-pack bands, 16,200 = the halving threshold, 16,900 = the SIL band, 20,959 = the downside runway) or an HTTP status code (403). Zero material external claims lack a SRC reference. | — (but see D-V7-3: three of those references resolve to rows that do not contain the figure) |
| REQ-CON-02 | PASS | I extracted the whole study to text and scanned for New South Wales, Queensland, South Australia, Western Australia, Tasmania, Northern Territory, NSW, QLD, WA and SA: **zero hits**. 47 occurrences of Victoria. Every jurisdiction-dependent obligation — worker screening, WWCC, WorkCover, portable LSL, payroll tax, disability worker registration, aged care screening — names Victoria or the Commonwealth. | — |
| REQ-CON-03 | PASS | I parsed the accessed dates myself: 74 rows, 53 at 2026-08-20 and 21 at 2026-09-07, against a currency date of 2026-09-07. None later than the currency date; the oldest is 18 days earlier, inside the 90-day window. The workbook README and the Help Hub now both state this distribution correctly. | — |
| REQ-CON-04 | **FAIL** | I ran my own scan over every Office file in `05_Outputs`, measuring explicit run sizes, inherited paragraph-style sizes and raw XML `sz=` attributes. Text: **the 10-point clause is genuinely met** — 0 sub-10pt runs in 3,223 study runs (1,587 of which inherit, and the three sub-10pt styles in the template are unused), minimum 10.0pt across 194 deck runs, 0 of 1,229 workbook cells below 10pt. Tables: 0 document tables above six columns in the .docx and .pptx; no defined Excel Table objects in the workbook. | The workbook's worksheet grids are **Scenarios 14 columns, Costs 7, TradeStudy3 7**. The requirement says "Every delivered Office artefact shall carry … zero tables of more than six columns" and its criterion is a flat threshold with no interpretive latitude. The count is 3, not 0. It reads as zero only under DEC-008, whose own record states "Interpreted by the project on 2026-09-07. **NOT RATIFIED BY ZAID**" and whose rationale field says a waiver cannot be self-granted. The project has correctly refused to self-grant the waiver and has correctly disclosed the widest worksheet in every run — and that is precisely why the requirement is not yet met. One decision from Zaid closes this; a builder cannot. |
| REQ-CON-05 | PASS | I parsed `03_Registers/ASM.csv`: 31 rows, zero with an empty Confidence field, zero with an empty 'What breaks if it is wrong' field. 26 Low, 5 Medium, matching the study's Table 12.1 caption. | — |
| REQ-CON-06 | PASS | I ran my own regex scan for email addresses, Australian mobile numbers, ABN-shaped identifiers, TFN patterns, API-key assignments, AWS keys and PEM private-key headers across 102 `.py/.md/.csv/.json/.html/.txt/.js` files in the controlled set. One hit, and it is a false positive: the digit run `06053495854` inside V6's own report, which is the tail of 1002.06053495854. Zero real hits. | — |
| REQ-MOE-01 | PASS | I read all thirteen rows of `03_Registers/ACT.csv`. ACT-010 now carries **2026-10-31** (or before the sixth client, whichever comes first) and ACT-012 carries **2026-12-19** (or before the tenth client). Every one of the thirteen carries an owner and a due date. The narrowest V6 failure is closed by the narrowest possible edit. | — (twelve of the thirteen are already past their stated due date as at 2026-09-07, which the criterion does not test) |
| REQ-MOP-01 | PASS | I re-implemented the unit extraction independently. Using the checker's own exemption list: 150 units carrying a dollar figure, 147 attributed, ratio **0.9800**. With **no exemptions at all**: 156 units, 150 attributed, ratio **0.9615**. Above 0.90 on either reading. The six unattributed units are three table captions and the three defect-history rows quoting what was wrong. | — |
| REQ-AC-01 | PASS | I rebuilt $1,002.06 from scratch: $30,000/12 = $2,500; less 10% (SRC-063) = $2,250; ÷ $103.11 (SRC-060) = 21.82135 hr; × ($43.03 × 1.1545) = $1,084.04 labour; + $250 pool; less $87.64 care management; less $318.75 service admin; less $7.50 bad debt = **$1,002.06**. And $435.30: 65 hr × $73.58 = $4,782.70; less 65 × $52.27576 = $3,397.92; less $949.48 = **$435.30**. Ratio 2.30200. Both figures are in study Table 15.3; workbook `AC_ClientEconomics!B17/C17` cache 1002.06053495854 and 435.300368749999, matching the model to the cent; C37 compares fifteen such figures. | — |
| REQ-AC-02 | PASS | Study Table 15.2 and workbook `AC_ClientEconomics` A22:D27 now span **$10,731 (Level 1, 1.80 hr/week, $894.25, $302.14)** to **$78,106 (Level 8, 13.11 hr/week, $6,508.83, $2,749.44)** — the actual bottom and top ongoing classifications, not a round $12,000. Every one of the 18 value cells is a live workbook formula over `A22:A27` and `AC_Inputs`; I opened them in formula view. | — |
| REQ-AC-03 | PASS | $606.50 / $7,278.00 / $22,578.00 appear in study Table 3 finding 6 and Table 17.1, in workbook `Structure!D5`, `D7 = D5*12` and `D8 = D5*36+D6`, and in `business_plan.json` at `value_proposition.benefits[3].target`, `areas_of_concern[4]` and `value_validation.conditions[4]`. All derive from `model_agedcare.structure_comparison()`. I recomputed 606.50 = 78+300+200+342/12 and 22,578 = 606.50×36 + 744. | — |
| REQ-AC-04 | PASS | Study Table 15.1a states the mechanism (dual care-management 15–20% plus package-management 10–15% replaced by a single 10% pool, landing on cost bases built for the old structure), cites the independent pass by name ("an independent verification pass was run specifically to resolve it (V4)"), and names the falsifier ("the DIRECT-SERVICES line turning negative in the quarterly sector report — not the total result"). I opened `V4_agedcare_margin_verification.md` and confirmed $16.10 is derived there from quoted text ($63.85 − $47.75), not asserted. | — (but see D-V7-3: the figure is still cited to SRC-068, which does not contain it, while the new SRC-074, which does, is cited nowhere) |
| REQ-BUS-01 | **FAIL** | The twin exists; I diffed it against a fresh rebuild and it is byte-identical in content. All twelve C-BP checks pass — I read all twelve and ran `validate()` myself. The A3 page and the twin are same-second current (both 07:06:08). I re-derived the entire CBA independently and it reconciles to the cent. | **The one-sheet property is produced by clipping, and the check that certifies it cannot fail.** `@media print` sets `html,body{height:296.5mm;overflow:hidden}` and `.sheet{height:296mm;overflow:hidden}`. Printed in Chromium with C34's exact call — `pg.pdf(prefer_css_page_size=True, print_background=True)`, honouring the file's own `@page{size:A3 landscape;margin:0}` — the output is 1 page at 420×297 mm **with sections 8 (partly), 9, the entire value-validation verdict block and the footer discarded**; I rasterised it and read the page. The cause is `@media(max-width:900px){.grid{columns:1}}` firing at print layout width, and `overflow:hidden` hiding the consequence. I then quadrupled the card content in a scratch copy and C34's call still returned exactly 1 page: **no amount of content can ever make this check fail.** At `format='A3', landscape=True` the page does render in full and legibly, so the fix is a layout that fits rather than a container that truncates — and a check with a negative test. |
| REQ-WEB-01 | PASS | `SYSTEM_BREAKDOWN.md` section 5 declares the slot with an explicit status — "**Status in this project: SLOT DECLARED, NOT BUILT**" — names GOV-B7.1/B7.2/B7.4 and states why it is not built. BKL-011 carries it in the backlog against REQ-WEB-01; COMPONENTS.csv scopes REQ-WEB-01 to C1. My own credential scan across every generated artefact returned zero hits. | — |

## New defects found

**D-V7-1. Every delivered artefact and every register is a projection of a source of truth that has since moved, and the status surface asserts a verification that no longer exists.**
`01_System/project_data.py` was last written at 07:07:36. Every register CSV was written at 07:05:55, every deliverable and every web surface at 07:06. Consequences I confirmed by opening the files: DEF-025 to DEF-034 exist in `project_data.DEF` (34 rows) and in **neither** `03_Registers/DEF.csv` **nor** `00_Handover/` (24 rows); the study's Appendix B2 publishes a 24-row defect history and states "24 defects, all closed"; the Help Hub says "DEF | Defects. 24 raised, 24 closed"; the README says "Defects | 24 raised, 24 closed". Worse, `03_Registers/REQ.csv` and `00_Handover/REQ.csv` both carry **26 Verified / 6 Open**, and `Dashboard.html` publishes "**26 / 32** Requirements INDEPENDENTLY verified — the verifier built none of them" and "Requirements: **26 of 32 verified** (81%)", while `project_data.IV_VERDICT` now derives **all 32 as Open and 26 as STALE**. The delivered study's Definition of Done gate 1 still reads "FAIL — 26 of 32 … Verdict source: `V3_closing_verdict.md` (independent verification pass 4, **2026-08-20**)". That 20-August-verdict-on-a-7-September-artefact is exactly the condition DEF-035 was raised to eliminate; the mechanism was built and the artefacts were never rebuilt through it. C40 is the only check that notices, and it names a byte count rather than the consequence.

**D-V7-2. The three-year net and the $90,000 sensitivity are wrong in the artefacts a reader actually opens.**
(a) ASM-029, in `project_data.py` and in both delivered ASM projections, states "a computed net of **$144,421.10**" while the computation now yields **$144,386.98**; `trade_study.py`'s docstring says the same stale figure two lines above the function that computes the right one. (b) ASM-028 states the $90k case as "$365,726.60 against the same **$420,147.70** of benefit, a net of **$54,421.10**"; the benefit is $420,113.58 and the net is $54,386.98. (c) Far more seriously, the delivered **study** (section 1, the ACT-013 box) and the delivered **dashboard** (ACT-013 row) both still say "At $90,000 the three-year return falls from **53 per cent to roughly 10**". The correct figures, which I computed twice, are **52.4% to 14.9%** — and ASM-028 already carries them. V6 raised this; the register was corrected and the two documents a human reads were not. The project describes ACT-013 as "the one input that changes the verdict", so this is the wrong number attached to the most decision-relevant sensitivity in the plan.

**D-V7-3. Three citations resolve to rows that do not contain the figure, and one figure has no source at all.**
(a) `$16.10 / $63.85 / $47.75` are cited to SRC-068 everywhere they are used (study finding 5, study Table 15.1a, RSK-011, deck slide 9); SRC-068's Value field is a document title with no figure in it. SRC-074 exists and carries all three and is cited nowhere. (b) `business_plan.json` and the A3 page state the NDIS market as "**Approximately 717,000** active participants nationally", cited to **SRC-001**, which is the *price schedule identity* ("v1.2, effective 1 July 2026"). No register anywhere contains 717,000 — the register figure is SRC-043's **774,456**. (c) The second market row, "Around 300,000 home-care recipients nationally", is cited to SRC-064, the classification-budget schedule, which contains budgets and no population. Check C-BP3 passes all three because it only tests that `size_source` is non-empty; C05 never reads the A3 or the JSON; C36 only tests that the identifier exists. This is DEF-031's defect class, unfixed and now present in three more places.

**D-V7-4. The DEF-032 fix left a false sentence behind it, and eight more false statements survive on the Help Hub, the dashboard and the README.**
`build_business_plan_json.py` line 347 formats `R['budget_sensitivity'][0][3]` into a hard-coded sentence: "the model brackets contribution from **$302.14** a client a month at a **$12,000** budget". `budget_sensitivity[0]` is now `(10731, 1.80, 894.25, 302.14)`. $302.14 is the contribution at **$10,731**, not at $12,000. The value moved when DEF-032 was fixed; the literal in the sentence did not. That false statement is in `business_plan.json`, on the A3 page and in the delivered PDF render. Alongside it, on surfaces DEF-029 was supposed to clean: the Help Hub says the decision pack has "**ten slides**" three times and the README twice — it has twelve; the deliverables map says the workbook has "**398 formulas, zero hard-coded results**" — I counted **455** formulas, and the workbook's own README now explicitly says two sheets *are* typed-in results, so the Help Hub directly contradicts the DEF-030 fix; the deliverables map and the "chart is empty" troubleshooting entry both say "**D1 to D10**" — there are twelve; the reading order says "study sections 1 and 9 → **the four ACT items**" when the study's own box names six; and the Help Hub calls the study "**32 pages**" in one entry and "**the 47-page study**" two entries later.

**D-V7-5. The published open-item count is short by two, and one of the two is the go/no-go decision itself.**
`project_data.parked_decisions()` matches the literal string `"Parked-Zaid"`. DEC-003 uses that string. **DEC-007** (the aged care/NDIS ordering — the decision the whole study escalates) and **DEC-008** (the six-column interpretation) are both recorded as `"Parked with Zaid"`, a different string, and are therefore invisible to `open_item_count()`. The dashboard prints "OPEN ITEMS … DEC parked with Zaid **1**" beside a decisions table that visibly lists three parked decisions. The published figure of 47 should be 49. This silently undoes DEF-015, whose fix note reads "A decision parked with Zaid is an open item … so it is now counted". C11 cannot catch it because the published count and the "count the registers actually produce" are the same function. Note that the DEF-028 fix is what created DEC-008, so a fix introduced an uncounted open item.

**D-V7-6. C22 does not do what its own description says, and the pack it certifies is stale.**
C22's description is "Every file in the transfer pack **re-derives from its source**: nothing missing, extra or altered", and REQ-SYS-17's statement is "The transfer pack shall re-derive from its sources". I read the function: it hashes each pack file and compares against the SHA-256 recorded in `MANIFEST.txt`, which was written at the same moment by the same generator. It proves the pack has not been tampered with since it was written. It re-derives nothing. That is why `00_Handover/REQ.csv` can say 26 Verified against a source of truth that says 0, and `00_Handover/ASM.csv` can carry the stale $144,421.10, while C22 reports zero discrepancies. This is the DEF-011 defect class ("checkers that assert more than they test") in a check that certifies a requirement.

**D-V7-7. Smaller things I would still record.** The T3 weights record — the GOV-B6.3 artefact — names criterion 5 "**Dependence** on the owner's own hours" while the scoring, the workbook and the study all use "**Independence** from"; the polarity of a fixed criterion is inverted in the only document that fixes it. Study Table 5.1 still cites "[Financial Model, **SupportCoordination sheet**]" for the $8,757 six-month runway; I searched every cell of that sheet and the figure is not in it. Workbook `AC_Inputs!F9` still says the Schedule F Level 3 rate is "used for care management and administration time" while `F23`/ASM-031, fourteen rows below, says the SACS rate is used instead — V6 raised this and it is unchanged. C12 still reads `if len(names) < 10: return False, "expected 10 diagrams"` and passes on `len(media) >= 10`, so two of the twelve diagrams could be deleted without failing it. The A3 footer still says "verified by PENDING — independent verification pass **V5** has not yet run against this plan"; V5 was the negative-test pass and the relevant pass is V7. The aged care critical path (Table 10.2) is the only critical path in the study with no diagram, while the NDIS one has D1 — and the aged care entity is the one the study recommends starting first. The CBA uses the *rounded* $1,002.06 for the aged care contribution while DEF-033 was specifically about not using the rounded $6.70 on the NDIS side; the effect is $0.18 over three years, but the principle is the one DEF-033 established. The README's section 1 still describes the project as being about "starting an NDIS provider business in Victoria" with no mention of aged care, and its reading order still points at Standard v3.7 while the governance line says v3.9 plus v4.0.

## What I could not verify, and why

- **Whether any external fact is true.** I verified internal consistency, derivation and citation, not the world. I did not open one of the 74 source URLs. Nothing in this report is evidence that $103.11, $73.58, $45.28, $43.03, $10,731, $78,106 or the July 2027 date are correct. Fifty-one of the 74 sources sit below High confidence with the reason stated, and that is the honest position, not a verified one.
- **The trade study scores themselves.** I re-derived every weighted total, the margin of 0.15 and all four flip points exactly. Whether option A's capital deserves a 3 and option C's a 5 is a judgement no verification pass can settle, and the project is right to escalate it. I also cannot verify the *ordering* claim behind trade study 3 — and neither can the project, which now says so in its own weights record.
- **Print behaviour on any engine but Chromium.** I measured with headless Chromium via Playwright, which is what C34 drives. I did not test Word, Safari, Firefox or a physical printer, and I did not open Chrome's interactive print preview. My REQ-BUS-01 failure rests on the browser the project's own check uses, printing under the rule the file itself declares.
- **Whether the Decision Pack renders as designed.** I read every run, table and picture shape through python-pptx and measured every font size, but I did not render the twelve slides, so I cannot speak to overflow, overlap or contrast on any of them.
- **The PDF rendition of the study.** `NDIS_and_Aged_Care_Business_Enabling_Study_v2.0.pdf` sits in `05_Outputs`, no requirement or checker addresses it, and I did not open it. I do not know whether it matches the .docx I verified.
- **Whether my own verdicts will survive transcription.** I proved in a sandbox that C31 and C41 pass regardless of what this file says. Nothing in the project mechanically connects the verdict map to this document, so the transcription step below is, today, an unverified builder statement — which is the thing GOV-C3.2 exists to forbid.

**Disclosure of my own footprint.** I ran the checker suite only inside a disposable copy of the tree in my scratch directory; `01_System/checker_run_log.txt` still carries the builder's 07:07:57 run untouched. Every Python process I started set `sys.dont_write_bytecode = True`; the `01_System/__pycache__` directory that C35 reports predates this pass (mtime 07:07:36, written alongside `project_data.py` by the builder's own run) and I did not add to it. `python 01_System/tidy.py` clears it. I created and modified exactly one file inside the project: this report.

## Closing verdict

**30 of 32 requirements PASS.**

The two failures are narrow and both are closable. REQ-CON-04 needs one decision from Zaid on DEC-008, or a restructured workbook. REQ-BUS-01 needs a layout that fits without `overflow:hidden` and a C34 with a negative test that proves it can fail. But read the new-defects section before reading the count as good news: nine of the eleven surfaces I opened are projections of a source of truth that moved after they were built, the delivered dashboard tells a stranger that 26 of 32 requirements are independently verified when none are, and the mechanism built to prevent exactly that can be defeated by writing a file with the right name.

## MACHINE-READABLE VERDICTS
```
REQ-SYS-01=PASS
REQ-SYS-02=PASS
REQ-SYS-03=PASS
REQ-SYS-04=PASS
REQ-SYS-05=PASS
REQ-SYS-06=PASS
REQ-SYS-07=PASS
REQ-SYS-08=PASS
REQ-SYS-09=PASS
REQ-SYS-10=PASS
REQ-SYS-11=PASS
REQ-SYS-12=PASS
REQ-SYS-13=PASS
REQ-SYS-14=PASS
REQ-SYS-15=PASS
REQ-SYS-16=PASS
REQ-SYS-17=PASS
REQ-SYS-18=PASS
REQ-CON-01=PASS
REQ-CON-02=PASS
REQ-CON-03=PASS
REQ-CON-04=FAIL
REQ-CON-05=PASS
REQ-CON-06=PASS
REQ-MOE-01=PASS
REQ-MOP-01=PASS
REQ-AC-01=PASS
REQ-AC-02=PASS
REQ-AC-03=PASS
REQ-AC-04=PASS
REQ-BUS-01=FAIL
REQ-WEB-01=PASS
```