arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2602.18571cs.SEcs.AI
收藏

Debug2Fix: 交互式调试能否帮助编码代理修复更多错误?

Debug2Fix: Can Interactive Debugging Help Coding Agents Fix More Bugs?

Microsoft(微软)

Spandan Garg, Yufan Huang

AI总结:

本文提出Debug2Fix框架,通过集成交互式调试子代理,提升编码代理在bug修复中的性能,实验显示在某些模型上性能提升超过20%,并证明更好的工具设计比切换更昂贵的模型更重要。

摘要:

尽管通过编码代理自动化软件开发的各个方面已取得显著进展,但其bug修复能力仍有很大提升空间。调试和运行时行为调查仍主要由开发者手动完成。流行的编码代理通常依赖静态代码分析或迭代测试-修复循环,类似于试错式调试。我们提出,开发者在调试代码时通常会访问丰富的运行时信息,而当前的设计限制使代理无法获取这些信息。尽管调试器在现代IDE和命令行工具中很普及,但它们却未被编码代理采用。本文介绍Debug2Fix,一种将交互式调试作为软件工程代理核心组件的新型框架,通过子代理架构整合Java和Python调试器,并在GitBug-Java和SWE-Bench-Live上评估,某些模型的性能相比基线提升了超过20%。此外,使用该框架,较弱的模型如GPT-5和Claude Haiku 4.5能匹配或超越更强的模型如Claude Sonnet 4.5,表明更好的工具设计往往比切换更昂贵的模型同样重要。最后,我们进行了系统性的消融实验,证明子代理架构和调试器整合的重要性。

英文摘要:

While significant progress has been made in automating various aspects of software development through coding agents, there is still significant room for improvement in their bug fixing capabilities. Debugging and investigation of runtime behavior remains largely a manual, developer-driven process. Popular coding agents typically rely on either static analysis of the code or iterative test-fix cycles, which is akin to trial and error debugging. We posit that there is a wealth of rich runtime information that developers routinely access while debugging code, which agents are currently deprived of due to design limitations. Despite how prevalent debuggers are in modern IDEs and command-line tools, they have surprisingly not made their way into coding agents. In this work, we introduce Debug2Fix, a novel framework that incorporates interactive debugging as a core component of a software engineering agent via a subagent architecture. We incorporate debuggers for Java and Python into our agent framework and evaluate against GitBug-Java and SWE-Bench-Live and achieve >20% improvement in performance compared to the baseline for certain models. Furthermore, using our framework, we're able to make weaker models like GPT-5 and Claude Haiku 4.5 match or exceed the performances of stronger models like Claude Sonnet 4.5, showing that better tool design is often just as important as switching to a more expensive model. Finally, we conduct systematic ablations demonstrating the importance of both the subagent architecture and debugger integration.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑