Verify a decision

{"actor":"judge","decided_at":"2026-08-25T18:02:55.762Z","decision":"review_revise","investigation_id":"ebfb4383-754a-4e4e-b2ae-1dbf8c1fccd0","new_score":0,"page_slug":"blockstream-jade-fake-firmware-phishing-campaign-august-2026","prev_score":5,"reason":"Recomputing disputed_pct from the review's own claim array (2 disputed + 2 unverifiable of 23 total, not the summary's mistaken 22) gives 17.4%, placing this page in the minor-issues revision band rather than approval. The most significant defect is claim_findings[5]: the page's title and lead frame an August 2026 'second wave' of Jade-specific phishing, but the two sources cited for it never mention Blockstream or Jade at all, only a sector-wide surge tied to a different vendor's exploit — an unsourced organizing premise that the page itself later hedges with the word 'alleged' in a later section. A second flaw, claim_findings[9], is a direct self-contradiction: the page describes a 2023 incident as 'physical mail' when both its own cited source and its own timeline entry describe an email scam. Weighing against these, the review found the page's core distinction — that Blockstream was impersonated, not compromised or shipping malicious firmware — held up consistently across the entire page, corroborated by multiple verbatim-verified primary-source quotes (claim_findings[0], [1], [3], [13], [19], [21]), and the remaining discrepancies (unattributed but accurate statistics, a Secure Boot citation mismatch, and one overstated superlative) are sourcing-hygiene problems rather than active misinformation. The reviewer's coverage gaps, none marked high priority, support treating this as a fixable sourcing and internal-consistency problem rather than a fundamentally unsound page.","score_delta":-10,"sequence_num":3,"submission_content_hash":null,"submission_id":null,"submission_kind":null,"submission_valence":null,"v":1}