OpenCollab：可编程协作与可控运行时的多智能体编码框架
原标题：OpenCollab: A Multi-Agent Coding Framework with Programmable Collaboration and Controllable Runtime
AI 摘要
该论文提出 OpenCollab，一个面向多智能体编码的可编程协作与可控运行时基础设施，通过统一抽象声明角色边界与通信拓扑，并在共享底座上执行不同配置以控制变量。研究提出 Adherence 指标量化声明组织与实际执行的一致性，发现仅改变单一维度即可使协作率从 47.2% 升至 97.2%，且审计的十个主流智能体框架均无法保证受控对比或验证行为交付。OpenCollab (Duo) 在 SWE-bench Pro、Terminal-Bench 2.1 和 DeepSWE 上取得新 SOTA，并显著优于单智能体基线。
正文节选
September 29, 2026 1 Introduction Autonomous coding agents have achieved remarkable progress, demonstrating promising capabilities to resolve complex issues in real-world software repositories (Jimenez et al., 2024; Wang et al., 2025). As real-world programming tasks grow in complexity, managing vast contexts and multi-stage execution within a single model becomes increasingly challenging. To push the boundaries of performance, the paradigm is shifting toward multi-agent systems (Li et al., 2023