Verify a decision

{"actor":"judge","decided_at":"2026-08-27T07:09:26.340Z","decision":"review_revise","investigation_id":"fdad1101-684b-4137-b3ab-e68d28005c64","new_score":53,"page_slug":"gho","prev_score":61,"reason":"Of 44 claims checked, only claim_findings[39] and claim_findings[42] are disputed and claim_findings[43] is unverifiable, giving a disputed_pct of 6.8% — well within the approve band, and the two disputed items are exact-day discrepancies of a few days on dates the surrounding prose already states more loosely (a repeg pinned to Feb 6 vs. the page's Feb 1, and an sGHO launch pinned to April 3 vs. the page's April 1), not disputes over the page's core claims. The reviewer also confirms the page gets the hard parts right: it classifies GHO's mechanism correctly throughout and reports the 2023-24 depeg accurately as both real and resolved. What moves this out of a clean approval are two high-priority coverage gaps and a concentrated staleness problem that the disputed_pct figure does not capture. Four stale findings (claim_findings[22-25]) show an entire 'Collateral Concentration' section, plus part of the GSM discussion, resting on a single LlamaRisk report from June 2024 — over two years old relative to the page's mid-2026 framing — presented with no as-of date, on exactly the topic (collateral risk) a reader would expect to be current. Separately, the review flags two high-priority coverage gaps: a complete absence of any mention of GHO's real, findable smart-contract audits (OpenZeppelin, SigmaPrime, ABDK, Certora), and no discussion of the GHO Stewards multisig that exercises the very governance discretion the page itself flags as a risk. Neither gap is a fabrication or an adversarial omission — the audit silence is arguably favorable-to-subject in effect since it withholds a credibility-boosting fact rather than distorting a risk one — but the combination of two high-priority gaps and a stale, undated core risk section is enough to warrant revision rather than a bare approval, per the reviewer's own coverage-gap guidance.","score_delta":-8,"sequence_num":3,"submission_content_hash":null,"submission_id":null,"submission_kind":null,"submission_valence":null,"v":1}