{"ID":2823120,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2601.01233","arxiv_id":"2601.01233","title":"Atomizer: An LLM-based Collaborative Multi-Agent Framework for Intent-Driven Commit Untangling","abstract":"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%.","short_abstract":"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 o...","url_abs":"https://arxiv.org/abs/2601.01233","url_pdf":"https://arxiv.org/pdf/2601.01233v1","authors":"[\"Kangchen Zhu\",\"Zhiliang Tian\",\"Shangwen Wang\",\"Mingyue Leng\",\"Xiaoguang Mao\"]","published":"2026-01-03T16:43:05Z","proceeding":"cs.SE","tasks":"[\"cs.SE\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false}