Verify a decision

{"actor":"judge","decided_at":"2026-08-25T03:49:02.077Z","decision":"review_revise","investigation_id":"d03ee4f0-598e-4137-a672-cb4ca8d53f45","new_score":42,"page_slug":"midnight-network-night-token","prev_score":52,"reason":"The fact-check found no disputed claims and confirmed 43 of 58 statements on the page, many independently beyond the page's own citations, which supports the page's core narrative. The 15.5% disputed_pct is driven entirely by 9 'unverifiable' items -- mostly unfalsifiable negatives (no audit found, no regulatory action found) and self-sourced project claims about its own token design -- rather than any claim shown to be false. However, the review identified a real editorial pattern worth correcting: on two separate figures for the July 2026 Wanchain bridge exploit (the dollar value stolen and the size of the price decline/rebound), the page silently adopted the highest or most favorable-to-narrative number from among its own cited sources without noting that those same sources disagree (claim_findings[34], claim_findings[39], claim_findings[56], claim_findings[57]). There is also a citation mismatch where a claim about 'integration costs' is attributed to an article that never discusses that topic (claim_findings[44]), and a present-tense claim that a named team member 'serves as' Head of Applied Cryptography that independent profiles suggest may be outdated (claim_findings[12]). Combined with two high-priority coverage gaps around on-chain forensic detail for the exploit and insider token concentration, these findings warrant revision rather than approval, but nothing in the review rises to the level of a disputed core allegation.","score_delta":-10,"sequence_num":3,"submission_content_hash":null,"submission_id":null,"submission_kind":null,"submission_valence":null,"v":1}