Search papers, labs, and topics across Lattice.
FeatX is a feature-oriented tool designed to streamline software evolution by allowing users to edit features rather than code directly. By extracting a hierarchical epic-feature structure and employing a three-stage Evolution Agent, it translates feature edits into code patches, significantly reducing cognitive load during the process. In user studies, FeatX demonstrated a 42.6% improvement in function-level modification localization F1 compared to leading LLM baselines, while also being more cost-effective at only $0.07 per operation.
Editing software through feature manipulation can enhance usability and efficiency, achieving a 42.6% boost in modification accuracy over traditional LLM approaches.
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/a496263365/FeatX/tree/demo, with a demonstration video at https://youtu.be/OZqKZ4Ii-yM.
