Verify a decision

{"actor":"judge","decided_at":"2026-08-25T12:04:49.952Z","decision":"review_revise","investigation_id":"1369ce3f-f679-4a86-8aa1-7d71e5e1ef03","new_score":57,"page_slug":"allo-protocol","prev_score":68,"reason":"Recomputing disputed_pct correctly from the reviewer's 23-entry claim array (the file's own summary block undercounts) gives (2 disputed + 4 unverifiable) / 23 = 26.1%, which falls in the upper half of the 10-30% band and warrants a revise, not an approve. The two disputed claims (claim_findings[16], claim_findings[17]) are order-of-magnitude errors on GTC's all-time high and low, contradicted by the page's own cited CoinMarketCap source -- serious but confined to a token-price section rather than a core allegation about fraud, security, or regulatory conduct, all of which the review left clean or only unresolved as negative claims. The four unverifiable findings, especially claim_findings[7] (the $67M/5,000-project impact figures and named partner list), reflect untraceable aggregate statistics rather than contradicted facts, and the reviewer separately flags a real mischaracterization at claim_findings[9], where organization-wide Gitcoin financials are misattributed specifically to the Grants Lab unit. Two high-priority coverage gaps reinforce the need for revision: sourcing the impact figures and correcting the Grants Lab financial attribution. Weighed against this, the page's disambiguation of Allo Protocol from the unrelated Allora Network and Allo (RWA) is confirmed accurate (claim_findings[21]), which keeps the penalty moderate rather than severe.","score_delta":-11,"sequence_num":3,"submission_content_hash":null,"submission_id":null,"submission_kind":null,"submission_valence":null,"v":1}