Verify a decision

{"actor":"judge","decided_at":"2026-08-26T10:07:33.427Z","decision":"review_revise","investigation_id":"083dd2eb-bd54-452d-884d-96a3b2872725","new_score":0,"page_slug":"quark-drainer","prev_score":0,"reason":"Recounting the reviewer's 67 claim findings (the '68' in its summary block is an accounting slip, not a truncated list) gives 44 confirmed, 11 partially_supported, 3 disputed, and 9 unverifiable, for a disputed_pct of 12/67 = 17.9% under the spec's (disputed+unverifiable)/total formula — squarely in the 10-30% revise band. The page's factual backbone (Quark Lab's identity, pricing model, chain/wallet coverage, and the Blockaid 'most prolific' designation) holds up well. The decisive issue is claim_findings[16]: the page tells readers Quark bypasses three named security tools, including Blockaid's own detection layer, and cites Blockaid's blog for this — but that article never makes the claim, the WalletGuard+Pocket Universe combination is documented elsewhere as an Angel Drainer capability, and even the vendor's own marketing only claims two of the three bypasses. Attributing a false defensive-tool-bypass claim to the tool vendor's own research is worse than an ordinary miscitation because a reader could reasonably conclude a working protection is compromised and stop using it; the reviewer's high-priority coverage gap on this exact point is well-founded and is treated as the primary driver of this verdict. Two further disputed items are real but lower-stakes: claim_findings[47] misdates Angel Drainer's shutdown as March 2024 when contemporaneous reporting (Forklog, CryptoTimes) places it in July 2024, and claim_findings[42] presents an inferred cause (\"improved detection and takedown of predecessor platforms\") for the 83% phishing-loss decline that the cited reporting instead attributes to market-cycle correlation and a shift to lower-value targets — inference dressed as sourced fact. A related item, claim_findings[43], cites an accurate $1.1B/212-incident figure but implies it reflects a rebound in drainer-specific theft when the source describes broader protocol exploits unrelated to wallet drainers; this was rated partially_supported rather than disputed and is a conflation/framing issue rather than a fabricated number. On the positive side, claim_findings[64] confirms the page's named-operator list (Security Alliance's Inferno/Rublevka/Eleven/Vanilla) is accurate and the page does not overreach into unsupported drainer-lineage claims, unlike sibling pages in this batch. Reviewer confidence (0.75) is reasonably high and there are no link_rot or stale-citation findings, so no additional penalty is warranted on those grounds, but the high-priority security-tool gap keeps the delta toward the upper end of the minor-issues band rather than the low end.","score_delta":-12,"sequence_num":3,"submission_content_hash":null,"submission_id":null,"submission_kind":null,"submission_valence":null,"v":1}