HL-40086
audit chainFields with a page citation
90.9%
computed fields excepted · 10 of 11 extractable fields
Bundle
0.12.0
stamped on this decision · INV-9
Citation coverage is below the 99% floor. `05 §3` is explicit that a decline here is a correctness incident rather than a reporting issue — a field nobody can trace is a defect, not a low-confidence value.
Down to the evidence
every recorded field · INV-1| field | value | step | resolves to | checker |
|---|---|---|---|---|
| BorrowerName | Nakamura, K. | File intake | application pg 1 | agreed |
| Occupancy | Primary residence | File intake | application pg 1 | agreed |
| CreditScore | 748 | Credit read | credit report pg 1 | agreed |
| HousingHistory | Owned, 6 yr | Credit read | credit report pg 2 | agreed |
| WageIncome | $9,240 / mo | Income orchestration | paystub pg 1 | agreed |
| SelfEmployment | None | Income orchestration | tax return pg 1 | not applicable |
| RentalIncome | $1,100 / mo | Income orchestration | Schedule E pg 3 | agreed |
| InvestmentAccounts | $38,050 | Income orchestration | stated by j.park | agreed |
| TotalObligations | $2,840 / mo | Obligations and DTI | credit report pg 3 | agreed |
| QualifyingPayment | $1,206.44 | Obligations and DTI | computed · qualifying_payment | agreed |
| DebtToIncomeRatio | 38.9% | Obligations and DTI | computed · dti | agreed |
| PropertyValue | $612,000 | Consumed collateral | appraisal pg 1 | agreed |
| LoanToValue | 70.4% | Consumed collateral | computed · ltv | agreed |
| FloodZone | Zone X | Insurance and flood | determination pg 1 | agreed |
InvestmentAccounts carries stated provenance because a person corrected it in blind review — the run had read a prior-quarter page and nothing flagged it. That is what makes this chain resolve to a person rather than only to a document.
Up to the authorisation
INV-9 · the chain a reviewer arrives fordecision HL-40086 · approve $95,000
└ bundle 0.12.0 · signed s.mendo
└ promoted by attempt Add title review, and show the checker the obligations
└ evidence 96.4% ±1.2, n 12
└ against baseline 92.4% ±2.0, n 5
This is the examination artifact. A reviewer who was not in the room can start from a decision, reach the page a value came from, and reach the experiment that authorised the agent which produced it — without asking anyone a question.