Adversarial Review Board v2

The 2026-06-05 second-pass adversarial review of the re-baselined Sooner underwriting (Closing Fee Financing, 60-month term, 5-year master-lease recourse). The first pass ran on a wrong product premise; this one is product-correct. Each issue shows severity, the skeptic's verdict, the reasoning, the fix, and its current disposition after the consolidated fix pass. Filter and search.

This is the issue catalog — not where you decide. The 36 open/planned findings here consolidate into 13 calls. To make decisions one-by-one, go to the Decision Register → (issue · fix · why it matters · your call). The full code-level specs live in the Resolution Contract →.
Run 52 agents Findings 63 raw · 60 catalogued Survived skeptic 14 / 14 high+critical Date 2026-06-05 Stage Pre-launch (legal vehicle settled)
v3 consolidation (2026-06-07). Every finding below has been dispositioned by the founder and resolved into the canonical resolution & implementation contract (docs/reviews/2026-06-07-underwriting-resolution-contract.md), grounded in the unit-economics model, the early-termination fee schedule, and the concentration policy. New artifacts shipped: the Expected Loss & Unit Economics model (HTML + Sooner EL Model.xlsx) and the Key Facts Statement (Sooner Key Facts Statement.docx). The loss numbers now reconcile: PD 4%/yr, net EL provision 1%/yr (PD × ~25% LGD), gross tail reserve 2.4%/yr. The economics were reframed by the founder's pushback: 53.4% ROE / 22.4% net IRR / 12.15% NIM via the OID. The employer recalibration is applied (ADNOC/SEHA/DHA → T1; Big Tech split). Remaining open items are the shared-contract extraction + engine parity (the canonical-page goal) and the rubric calibration once rental data is sourced.

All identified issues

Severity
Dimension
Status

Fix backlog (post-consolidation)

Sequenced most-impactful first. Items struck/green are addressed in the consolidated pass. Effort: S ≤ 1 day · M ≤ 1 week · L > 1 week or cross-team.

Cross-cutting themes

What the latest research says we're getting wrong