一句话：提出特性导向的 LLM 代码演化工具 FeatX，将特性编辑自动映射并翻译为仓库级代码补丁。
现有基于大语言模型的软件演化交互范式以代码为中心，需要人工管理上下文并反复迭代提示词。本文提出 FeatX，一种“通过编辑特性来编辑软件”的特性导向工具。给定已有代码仓库，FeatX 先抽取分层的 epic-feature 结构，并建立显式的特性到代码映射；随后调用三阶段 Evolution Agent，将用户的特性编辑意图翻译为代码补丁。整个工作流通过四个协同面板呈现给用户。作者在受控用户研究与 38 个真实特性编辑提交的重放实验上进行评估，结果表明 FeatX 相比原生 ChatGPT 显著降低认知负荷并提升可用性；在函数级修改定位 F1 上，相对强 LLM 基线取得 42.6% 的相对提升，且总成本仅约 0.07 美元。
Large language models (LLMs) are increasingly used for software evolution, yet most interaction paradigms remain code-centric and require manual context management and prompt iteration. We present FeatX, a feature-oriented tool for editing software by editing features. Given an existing repository, FeatX extracts a hierarchical epic-feature structure with explicit feature-to-code mappings, then invokes a three-stage Evolution Agent to translate feature edits into code patches. The workflow is exposed through four coordinated panels. Across a controlled user study and replay experiments on 38 real-world feature-editing commits, FeatX significantly reduces cognitive load and improves usability compared with vanilla ChatGPT. It also achieves a 42.6% relative improvement in function-level modification localization F1 over strong LLM baselines, at substantially lower cost ($0.07 in total). The tool and collected dataset are available at https://github.com/PKU-SoftwareReuse/FeatX, with a demonstration video at https://youtu.be/OZqKZ4Ii-yM.
@inproceedings{abs-2606-31206,
title = {FeatX: Editing Software by Editing Features for Repository-Level Code Evolution},
author = {Xutian Li and Yifeng Zhu and Xianlin Zhao and Yanzhen Zou and Lu Zhang and Bing Xie},
booktitle = {ASE},
year = {2026},
doi = {10.1145/3832783.3834646}
}