Problem
Research questions and friction points this paper is trying to address.
agent-authored pull requests
code review
human-agent collaboration
coordination dynamics
software development
Humans Integrate, Agents Fix: How Agent-Authored Pull Requests Are Referenced in Practice
🤖 AI Summary
This study investigates the collaborative patterns and citation intents between human developers and AI coding agents in code reviews centered on agent-generated pull requests (PRs). Leveraging the AIDev dataset, it introduces the first taxonomy for classifying citation intent in agent-generated PRs and conducts an empirical analysis of PR citation relationships in large-scale open-source projects through data mining and content analysis. The findings reveal a clear division of labor: humans predominantly cite PRs to integrate new features, whereas AI agents focus on bug fixes. Moreover, PRs involved in citation networks—either citing or being cited—exhibit significantly longer lifecycles and review durations than isolated PRs, indicating higher coordination costs. The work further identifies the emergence of a “meta-collaboration” workflow and outlines three directions for optimizing AI agent integration in code review processes.
📝 Abstract
Although coding agents have introduced new coordination dynamics in collaborative software development, detailed interactions in practice remain underexplored, especially for the code review process. In this study, we mine agent-authored PR references from the AIDev dataset and introduce a taxonomy to characterize the intent of these references across Human-to-Agent and Agent-to-Agent interactions in the form of Pull Requests (i.e. PRs). Our analysis shows that while humans initiate most references to agent-authored PRs, a substantial portion of these interactions are AI-assisted, indicating the emergence of meta-collaborative workflows, where humans mostly use references to build new features, whereas agents make them to fix errors. We further find that referencing/referenced PRs are associated with substantially longer lifespans and review times compared to isolated PRs, suggesting higher coordination or integration effort. We then list three key takeaways as potential future research directions into how to utilize these dynamics for optimizing AI coding agents in the code review process.
Innovation
Methods, ideas, or system contributions that make the work stand out.
agent-authored pull requests
code review
human-AI collaboration
coordination dynamics
meta-collaborative workflows
🔎 Similar Papers
No similar papers found.
