[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-study-finds-multi-agent-coding-bots-often-ignore-their-own-rules":10,"sections":34},{"siteName":4,"siteTagline":5,"publisherName":4,"contactEmail":6},"The Revision","Tech news, decoded.","editor@therevision.news",{"gaMeasurementId":8,"adsenseClientId":9},"G-ZW2MV82GYR","ca-pub-8533917693782264",{"article":11},{"id":12,"slug":13,"title":14,"dek":15,"body_md":16,"tags_json":17,"published_at":18,"created_at":19,"updated_at":20,"status":21,"review_note":22,"review_notes":23,"image_url":22,"persona_id":22,"persona_name":22,"section":24,"tags":25,"sources":29,"feedback":33,"feedback_at":22,"cost_usd":33,"total_tokens":33},9005,"study-finds-multi-agent-coding-bots-often-ignore-their-own-rules","Study Finds Multi-Agent Coding Bots Often Ignore Their Own Rules","A new framework shows AI coding agents often abandon their assigned team roles, with adherence to design ranging from 47% to 97% depending on setup.","A new open research framework exposes a basic problem with AI coding teams: they don't always follow the org chart they're given.\n\nOpenCollab is a system for running multi-agent coding setups on a shared, controllable runtime. It tracks every action through fine-grained event streams and introduces a metric called Adherence, which measures whether agents actually stick to their assigned collaboration structure instead of drifting into ad-hoc behavior. The researchers found adherence varies wildly - from 47.2% to 97.2% - just by changing a single configuration dimension. A two-coder workflow built on OpenCollab also beat established harnesses including Mini-SWE-agent, Codex CLI, and Claude Code on coding benchmarks, while OpenCollab's single-agent setup used fewer tokens than any other configuration tested.\n\nThat adherence gap matters because most multi-agent coding research assumes the organization being tested is the organization actually running. If agents quietly ignore their assigned roles, performance comparisons between frameworks are really comparing unknown, uncontrolled behavior rather than the designs on paper. OpenCollab's event tracking gives researchers a way to isolate whether a result comes from clever coordination or from agents improvising in a way that happened to work.\n\nWorth remembering next time a startup pitches its multi-agent product: plenty of these systems look like an org chart on a slide and behave like a group chat in production.","[\"ai\",\"coding-agents\",\"multi-agent-systems\",\"research\"]","2026-10-01T04:00:00.000Z","2026-10-01T15:20:08.862Z","2026-10-01T15:20:11.739Z","published",null,[],"ai",[24,26,27,28],"coding-agents","multi-agent-systems","research",[30],{"name":31,"url":32},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2609.38345",0,{"sections":35},[36,39,43,48,53,58,62,67,72,76,81,86,91,96],{"name":37,"slug":24,"count":38,"latest_published_at":18},"AI",5487,{"name":40,"slug":41,"count":42,"latest_published_at":18},"Security","security",809,{"name":44,"slug":45,"count":46,"latest_published_at":47},"Policy","policy",429,"2026-10-01T02:26:17.000Z",{"name":49,"slug":50,"count":51,"latest_published_at":52},"Deals","deals",298,"2026-09-30T21:00:26.000Z",{"name":54,"slug":55,"count":56,"latest_published_at":57},"Hardware","hardware",196,"2026-09-30T13:00:00.000Z",{"name":59,"slug":60,"count":61,"latest_published_at":18},"Science","science",162,{"name":63,"slug":64,"count":65,"latest_published_at":66},"Consumer Tech","consumer-tech",149,"2026-09-30T22:57:11.000Z",{"name":68,"slug":69,"count":70,"latest_published_at":71},"Dev Tools","dev-tools",93,"2026-10-01T02:30:48.000Z",{"name":73,"slug":74,"count":70,"latest_published_at":75},"Software","software","2026-09-30T21:41:11.000Z",{"name":77,"slug":78,"count":79,"latest_published_at":80},"Startups","startups",84,"2026-09-30T20:39:09.000Z",{"name":82,"slug":83,"count":84,"latest_published_at":85},"Gaming","gaming",51,"2026-09-30T16:24:30.000Z",{"name":87,"slug":88,"count":89,"latest_published_at":90},"General","general",50,"2026-09-30T21:37:54.000Z",{"name":92,"slug":93,"count":94,"latest_published_at":95},"Reviews","reviews",31,"2026-09-28T14:31:34.000Z",{"name":97,"slug":98,"count":99,"latest_published_at":100},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]