Robert Stowe

How core and new contributors use coding agents differently

A white catamaran moored at a harbour wall in front of a row of pastel buildings, under a bright sky with scattered clouds
Are We All Using Agents the Same Way? An Empirical Study of Core and Peripheral Developers Use of Coding Agents
Autonomous AI agents are transforming software development and redefining how developers collaborate with AI. Prior research shows that the adoption and use of AI-powered tools differ between core and peripheral developers. However, it remains unclear how this dynamic unfolds in the emerging era of autonomous coding agents. In this paper, we present the first empirical study of 9,427 agentic PRs, examining how core and peripheral developers use, review, modify, and verify agent-generated contributions prior to acceptance. Through a mix of qualitative and quantitative analysis, we make four key contributions. First, a subset of peripheral developers use agents more often, delegating tasks evenly across bug fixing, feature addition, documentation, and testing. In contrast, core developers focus more on documentation and testing, yet their agentic PRs are frequently merged into the main/master branch. Second, core developers engage slightly more in review discussions than peripheral developers, and both groups focus on evolvability issues. Third, agentic PRs are less likely to be modified, but when they are, both groups commonly perform refactoring. Finally, peripheral developers are more likely to merge without running CI checks, whereas core developers more consistently require passing verification before acceptance. Our analysis offers a comprehensive view of how developer experience shapes integration offer insights for both peripheral and core developers on how to effectively collaborate with coding agents.
arxiv.org

A new study on arXiv looked at 9,427 pull requests opened with coding agents and asked whether the people at the core of a project use agents the same way as the people at its edges.

They do not. Occasional contributors use agents more often and for everything: bug fixes, features, documentation, tests. Core developers point them mostly at documentation and testing, and their agentic pull requests are merged more often. The finding I keep thinking about: peripheral developers are more likely to merge without running the continuous integration (CI) checks, while core developers more consistently require passing verification first.

I think that is a review policy finding. The people with the least context about a codebase are the most likely to let an agent do more and skip the check at the end, which is what you would expect when the tool is easy and the code is unfamiliar.

My view, after years of running teams that mix long-tenured engineers with new ones, is that one review policy for everyone is the wrong shape once agents are involved. Asking more verification of the newcomer's pull request is not a judgment on the newcomer. It is a judgment on where the risk sits.

Does your team's review process treat an agent-written pull request differently depending on who opened it?

Photo source: https://photos.robertstowe.com/bermuda