Verify a decision

{"actor":"judge","decided_at":"2026-08-18T20:31:16.494Z","decision":"review_revise","investigation_id":"2f6eca6d-bddc-4eeb-9bac-ad83f94cb1b8","new_score":58,"page_slug":"zetachain","prev_score":42,"reason":"The reviewer's incident classification is correct and the current score of 42 (WARNING) does not match the entity's conduct under the rubric. The page's flagship incident, the $334K GatewayEVM exploit, is a suffered hack against ZetaChain's own operational wallets by an external, premeditated attacker (calibration_assessment.incident_attributions[0], attribution 'b') — no external user funds were lost, the team paused operations within minutes, disclosed promptly, and shipped a patch. Per the rubric, a suffered primary incident caps the band at CAUTIONARY (50-69) regardless of the negligent bug-bounty triage failure, which is culpable but does not amount to own-fraud. The other flagged items — the disclosed sybil-reward airdrop policy and TSS/validator-concentration risk (attributions 'a') — are legitimate but minor own-conduct concerns, not fraud, and correctly argue against a top-of-range score rather than against CAUTIONARY itself. The PCRisk/ScamAdviser phishing-site findings (attribution 'c') describe brand impersonation by unrelated third parties and should carry near-zero weight under the anti-conflation rule; nothing in the current 42 appears explainable by that section carrying disproportionate negative weight, but it should not be allowed to anchor the score down further. A mid-CAUTIONARY score of 58 appropriately reflects: no user losses, rapid remediation, and prompt disclosure (positive), weighed against real negligence in dismissing the prior bug report, ongoing token-price deterioration now worse than the page states, and an unaddressed business-model pivot away from cross-chain bridging that signals continuing uncertainty (negative).","score_delta":16,"sequence_num":3,"submission_content_hash":null,"submission_id":null,"submission_kind":null,"submission_valence":null,"v":1}