Kan Watanabe, Rikuto Tsuchida, Takahiro Monno, Bin Huang, Kazuma Yamasaki, Youmei Fan, Kazumasa Shimari, Kenichi Matsumoto
This study examines how AI coding agents' pull request descriptions differ and how these differences affect human reviewers' responses and merge outcomes on GitHub.
Researchers investigated how AI coding agents, which use large language models to generate pull requests on GitHub, differ in their communication styles. They analyzed how the structure and characteristics of these pull requests influence human reviewers' engagement, response times, and decisions to merge the code. The study found that different AI agents have unique styles that lead to varying levels of interaction and success in getting their code merged. This suggests that the way AI presents its work can significantly impact collaborative software development with humans.
