一句话：提出用 GitHub 贡献数据对开发者不活跃进行时间预测，可在 7 天窗口前提前预警 65–75%。
开源软件项目依赖核心开发者的持续参与，因此当贡献者长期离开或脱离时，项目会变得脆弱。已有工作刻画了不活跃和开发者休息现象，但维护者仍缺少实用工具来提前预判。本文将开发者脱离建模为一个针对 GitHub 用户贡献的时间预测问题，并在开源项目上进行了初步评估。结果显示，该方法能在不活跃状态转变前的 7 天窗口中正确标记 65–75% 的情况，表明 GitHub 数据可以支持面向开源开发者休息的早期预警系统。该研究为自动预警系统奠定了基础，帮助维护者预见并缓解项目风险，也推动了对开源社区开发者可持续性的更广泛理解。
Open Source Software (OSS) projects rely on the sustained participation of core developers, making them vulnerable when contributors take extended breaks or disengage. Prior work has characterized inactivity and developer breaks, yet maintainers lack practical tools to anticipate them. We model developer disengagement as a temporal prediction problem against GitHub user contributions. Preliminary results on OSS projects show that our approach correctly flags 65–75% of 7-day windows preceding inactivity transitions, suggesting that GitHub data can support early-warning systems for OSS developers’ breaks. This research contributes a foundation for automated early-warning systems that help maintainers anticipate and mitigate project risks, representing a step toward a broader understanding of developer sustainability in OSS communities.