生成与验证：安全关键汽车软件工程中的人机协作
原标题：Generation and Verification: Human-Agent Collaboration for Safety-Critical Automotive Software Engineering
AI 摘要
该论文提出面向安全关键汽车软件的编码助手，基于模型上下文协议（MCP）向开发者暴露生成、编辑、验证三个工具，使开发者可随时对任意代码主动运行验证，而非依赖智能体自行检查。研究使用AUMOVIO生产汽车代码构建数据集，将每个函数与需求文本和UML行为配对，并比较提示工程与领域微调两种定制策略。在为期两周的生产项目试点中，开发者反馈验证是最常用工具，75%偏好更快的提示工程配置；MISRA和AUTOSAR风格的模式检查显示，提示工程在C语言上违规更少，微调在C++架构模式符合性上更好。
正文节选
Generation and Verification: Human-Agent Collaboration for Safety-Critical Automotive Software Engineering Abstract Research on coding agents mainly focuses on generation quality. Yet, safety-critical automotive software makes two additional demands. It prioritizes verification, since every change is reviewed, traced to a requirement, and checked against industry standards such as MISRA and AUTOSAR before it can be committed. It is also a narrow domain, written against layered architectures, emb