SWE-chat: Coding Agent Interactions From Real Users in the Wild
2026-04-22 • Artificial Intelligence
Artificial IntelligenceComputers and SocietySoftware Engineering
AI summaryⓘ
The authors created SWE-chat, a large dataset with 6,000 real coding sessions where developers interact with AI coding agents. They found that in many sessions, AI writes most of the code, but sometimes humans do it all. The AI code is only partly useful since less than half of it is kept, and it can introduce more security issues than human-written code. Users often correct or reject AI suggestions, showing current AI tools have notable flaws in real use. SWE-chat helps researchers understand how AI coding agents actually perform outside of controlled tests.
AI coding agentsdatasetsoftware engineeringcode completionsecurity vulnerabilitieshuman-computer interactionprogram synthesisdeveloper workflowsempirical studycode quality
Authors
Joachim Baumann, Vishakh Padmakumar, Xiang Li, John Yang, Diyi Yang, Sanmi Koyejo
Abstract
AI coding agents are being adopted at scale, yet we lack empirical evidence on how people actually use them and how much of their output is useful in practice. We present SWE-chat, the first large-scale dataset of real coding agent sessions collected from open-source developers in the wild. The dataset currently contains 6,000 sessions, comprising more than 63,000 user prompts and 355,000 agent tool calls. SWE-chat is a living dataset; our collection pipeline automatically and continually discovers and processes sessions from public repositories. Leveraging SWE-chat, we provide an initial empirical characterization of real-world coding agent usage and failure modes. We find that coding patterns are bimodal: in 41% of sessions, agents author virtually all committed code ("vibe coding"), while in 23%, humans write all code themselves. Despite rapidly improving capabilities, coding agents remain inefficient in natural settings. Just 44% of all agent-produced code survives into user commits, and agent-written code introduces more security vulnerabilities than code authored by humans. Furthermore, users push back against agent outputs -- through corrections, failure reports, and interruptions -- in 44% of all turns. By capturing complete interaction traces with human vs. agent code authorship attribution, SWE-chat provides an empirical foundation for moving beyond curated benchmarks towards an evidence-based understanding of how AI agents perform in real developer workflows.