跳到正文
arXiv cs.AI· Jonathan Ullrich, Julien Siebert, Michael Kl\"as, Adam Trendowicz, Andreas Jedlitschka·· 4 小时前AI 评分22

用智能体轨迹作为软件质量保证的证据

Agent Trajectories as Evidence for Software Quality Assurance

AI 导读

针对自主软件智能体在有限人工监督下做出的隐含假设难以被最终代码和通过的测试暴露,研究者提出将可观测的智能体轨迹与软件及智能体规格、AI 生成产物一并分析,作为定向质量保证的证据,用于筛选需要重点审查的关键决策。该方法源于工业实践经验,现有执行记录与初步分析方法使其具备可行性,但轨迹本身无法证明决策合理或纠正措施有效。

正文 · 原文

View PDF HTML (experimental)

Abstract:Autonomous software agents increasingly plan, implement, and test software with limited human oversight. Their decisions may introduce implicit assumptions that final code and passing tests do not adequately expose. We propose analyzing observable agent trajectories alongside software and agent specifications and AI generated artifacts as evidence for targeted quality assurance. The aim is to prioritize consequential decisions for scrutiny, investigate their possible contribution to software problems, and inform improvements to specifications and development practices. Industrial experience motivates the idea; available execution records and preliminary analysis methods make it plausible. However, trajectories alone do not establish that decisions are justified, outcomes acceptable, or corrective changes effective. We ask how this evidence can support review, diagnosis, and continuous quality improvement, and when simpler QA approaches suffice. Perspectives from requirements engineering, verification and validation, agent systems, causal inference, empirical software engineering, and human factors could help develop and evaluate the idea.
Comments: 3 pages, submitted to ICSE2027 AMPLIFY Workshop this https URL
Subjects: Software Engineering (cs.SE); Artificial Intelligence (cs.AI)
ACM classes: D.2.0
Cite as: arXiv:2611.22323 [cs.SE]
  (or arXiv:2611.22323v1 [cs.SE] for this version)
  https://doi.org/10.48550/arXiv.2611.22323

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Julien Siebert [view email]
[v1] Tue, 17 Nov 2026 07:22:29 UTC (84 KB)

来源:arXiv cs.AI · arxiv.org