Verify a decision

{"actor":"judge","decided_at":"2026-06-09T20:52:25.392Z","decision":"review_approve","investigation_id":"4a96af88-86dd-43ea-aa9a-acfba1b7ebc3","new_score":9,"page_slug":"rhea-finance-exploit-april-2026","prev_score":12,"reason":"The review confirmed 18 of 25 claims outright and found no actively disputed claims, placing the disputed rate at 4% — well within the approve threshold. All core factual claims about the exploit itself (loss amounts, attack mechanism, attacker wallet prefix, Tether freeze, and recovery figures) are confirmed by multiple independent sources including Tier-1 outlets. The five partially supported claims are confined to peripheral comparisons and background context: claim_findings[13] overstates the KyberSwap 2023 loss figure ($54.7M stated vs approximately $47–49M per Hacken and KyberSwap's own post-mortem) and imprecisely characterizes the vulnerability similarity; claim_findings[14] understates the Mango Markets loss by $2M ($114M stated vs $116M widely reported); claim_findings[3] cites an unverifiable 95% TVL share figure with no Tier-1 source. These are errors in contextual comparisons, not in the primary event reporting. A minor score adjustment is applied to flag the KyberSwap figure as a correctable inaccuracy. Two high-priority coverage gaps (on-chain primary verification, official post-mortem direct link) should be addressed in a future revision but do not affect approval.","score_delta":-3,"sequence_num":3,"submission_content_hash":null,"submission_id":null,"submission_kind":null,"submission_valence":null,"v":1}