Verify a decision

{"actor":"judge","decided_at":"2026-08-25T07:05:01.163Z","decision":"review_revise","investigation_id":"98b78ed6-e47d-48f6-99c6-5d091e26f8cc","new_score":54,"page_slug":"polkadot-dot","prev_score":62,"reason":"Most of this page checks out: 27 of 38 fact-checked claims (71%) were independently confirmed, and the raw disputed rate (claim_findings summary, disputed_pct 0.0789) is low. However, two of the non-confirmed findings sit inside the page's own highest-severity sections and are more than minor imprecision. claim_findings[24] states the treasury posted nearly 101,000 DOT in negative net flows in Q4 2025, but the page's own cited treasury report says the opposite — a net profit, described as the first positive quarterly result since OpenGov governance began. claim_findings[18] gives the Hyperbridge exploit loss as $237,000–$250,000; that was accurate on the day, but Hyperbridge itself later revised the loss to roughly $2.5 million, about ten times higher, and the page does not disclose the revision — compounded by claim_findings[20], which describes the exploit as contained to the Ethereum side when losses actually spanned four chains. A smaller, lower-stakes discrepancy also appears at claim_findings[30] (a cited 3.6% price move versus 3.02% in the page's own source) and claim_findings[13] (naming one ETF custodian where reporting describes three). Because the two most consequential errors both understate or reverse negative outcomes in the sections the page itself labels high severity (the treasury and exploit sections), rather than being scattered minor date or wording slips, we are treating this page as needing correction rather than a clean pass, with a modest score penalty reflecting that the overall claim base is still largely sound.","score_delta":-8,"sequence_num":3,"submission_content_hash":null,"submission_id":null,"submission_kind":null,"submission_valence":null,"v":1}