Verify a decision

{"actor":"judge","decided_at":"2026-08-18T16:55:18.552Z","decision":"review_revise","investigation_id":"3ff37634-d276-4760-a4b9-2c4b57a50a6d","new_score":60,"page_slug":"dodo-amm","prev_score":42,"reason":"The reviewer classifies the sole material incident on the page — the March 8, 2021 V2 Crowdpooling exploit — as attribution 'b' (suffered), noting DODO was the victim of an external smart-contract attack, published a transparent post-mortem, recovered roughly 82% of the drained funds within 24 hours (claim_findings[3]-[4], both confirmed), and has had no further exploits, fraud allegations, or regulatory actions in over five years of subsequent operation (calibration_assessment.incident_attributions[0]). Per the scoring rubric's own worked example, a suffered hack with this recovery profile caps exposure at CAUTIONARY (50-69), not WARNING (20-49); the current 42/WARNING score appears to be treating the incident as entity-attributable risk rather than victimhood. The reviewer's recommended 60 appropriately sits mid-band rather than at the top, reflecting two legitimate residual drags that are unrelated to fraud: the exploited Crowdpooling contracts shipped without a pre-launch audit (a process gap, not misconduct), and the protocol's TVL/market relevance has declined sharply since 2021. This is a correction in the entity's favor — the page's own hedged language ('an attacker exploited,' 'voluntary restitution') already reflects the suffered-hack read; the score band did not.","score_delta":18,"sequence_num":3,"submission_content_hash":null,"submission_id":null,"submission_kind":null,"submission_valence":null,"v":1}