一句话：提出 Atomizer，用意图导向思维链与分组者-审查者多智能体协作循环，实现复合提交的语义意图解缠。
复合提交常将多个无关关注点纠缠在一起，严重影响程序理解与维护。现有自动解缠方法（尤其是基于图聚类的先进方法）存在两个根本局限：过度依赖结构信息而无法把握变更背后的语义意图；采用“单遍”算法，缺乏人类审查中关键的反思与精化机制。为此，本文提出 Atomizer——一个面向复合提交解缠的协作多智能体框架。针对语义缺失，Atomizer 采用意图导向思维链（IO-CoT）策略，引导大语言模型结合代码结构与语义信息推断每次代码变更的意图。针对单遍分组的不足，框架设置分组者与审查者两个智能体，建立协作精化循环，迭代改进分组，直至同一簇内所有变更共享相同的底层语义意图。在 C# 和 Java 两个基准数据集上的大量实验表明，Atomizer 显著优于多个代表性基线，平均分别超越最先进的图方法 6.0% 和 5.5% 以上；在复杂提交上优势更加明显，超过 16%。
Composite commits, which entangle multiple unrelated concerns, are prevalent in software development and significantly hinder program comprehension and maintenance. Existing automated untangling methods, particularly state-of-the-art graph clustering-based approaches, are fundamentally limited by two issues. (1) They over-rely on structural information, failing to grasp the crucial semantic intent behind changes, and (2) they operate as “single-pass” algorithms, lacking a mechanism for the critical reflection and refinement inherent in human review processes. To overcome these challenges, we introduce Atomizer, a novel collaborative multi-agent framework for composite commit untangling. To address the semantic deficit, Atomizer employs an Intent-Oriented Chain-of-Thought (IO-CoT) strategy, which prompts large language models (LLMs) to infer the intent of each code change according to both the structure and the semantic information of code. To overcome the limitations of “single-pass” grouping, we employ two agents to establish a grouper-reviewer collaborative refinement loop, which mirrors human review practices by iteratively refining groupings until all changes in a cluster share the same underlying semantic intent. Extensive experiments on two benchmark C# and Java datasets demonstrate that Atomizer significantly outperforms several representative baselines. On average, it surpasses the state-of-the-art graph-based methods by over 6.0% on the C# dataset and 5.5% on the Java dataset. This superiority is particularly pronounced on complex commits, where Atomizer’s performance advantage widens to over 16%.
@inproceedings{paperbot11932,
title = {Atomizer: An LLM-based Collaborative Multi-Agent Framework for Intent-Driven Commit Untangling},
author = {Kangchen Zhu and Zhiliang Tian and Shangwen Wang and Mingyue Leng and Xiaoguang Mao},
booktitle = {ICSE},
year = {2026},
doi = {10.1145/3744916.3787773}
}