一句话：提出并部署 FlowAgent，在 Google 提交前 CI 流程中以低延迟 LLM 智能体自动修复测试失败，人工评估修复准确率达 67.18%。
手工修复程序失败耗时且打断开发者，尤其在持续集成系统的提交前阶段；现有基于大语言模型的自动程序修复主要面向提交后离线工作流，无法满足开发者切换上下文前实时修复所需的低延迟要求。本文提出 FlowAgent，一个部署在 Google 内部、用于提交前 outer-loop 工作流的自动修复测试失败的 AI 智能体。它集成到 Critique 与 Cider 开发者工具中，采用 ReAct 风格的生成-验证循环，并设置严格的执行前与执行后弃权过滤器，在严格延迟约束下保证建议质量。评估方面，对 195 个真实测试失败的人工评估显示修复准确率为 67.18%；全公司部署后 FlowAgent 在 295,508 个 change 上给出修复建议，其中 65,069 个被预览、28,554 个被采纳，访谈反馈表明开发者认可其正确修复能力及将自主修复智能体集成到工业软件工程流程中的做法。
Manual repair of program failures is time-consuming and disruptive for software developers, particularly during the pre-submit phase where test failures occur within continuous integration systems. While Automated Program Repair has seen significant advancement through Large Language Models, existing state-of-the-art techniques primarily focus on post-submit workflows, operating offline without the low-latency requirements necessary to assist developers in real-time within their flow before they switch context. In this paper, we introduce FlowAgent, an AI agent deployed at Google to automatically repair test failures in the pre-submit outer-loop workflow inside continuous integration systems. Integrated into Google’s internal developer tools, Critique and Cider, FlowAgent utilizes a ReAct-style generate-and-validate loop, as well as rigorous pre-execution and post-execution abstention filters to ensure high-quality suggestions under strict latency constraints. Based on our case studies, FlowAgent is highly effective. First, a manual evaluation conducted on 195 real-world test failures demonstrated 67.18% accuracy in suggesting correct fixes. Following its Google-wide deployment, FlowAgent suggested fixes on 295,508 changes, of which developers previewed 65,069 and applied 28,554. Developer feedback from interviews indicate that the agent is useful in suggesting correct fixes, integration of autonomous repair agents into industrial software engineering workflows is received well, while interesting challenges and opportunities still remain.
@inproceedings{paperbot12415,
title = {Catching Developers in the Flow: Low-Latency Agentic Program Repair at Google Scale},
author = {Celal Ziftci and Spencer Greene and Ray Liu and Livio Dalloro and Lorenzo Dini},
booktitle = {ASE},
year = {2026},
doi = {10.1145/3832783.3834511}
}