一句话:首届 FSE-AIware 竞赛为基于智能体的 Python 依赖解析提供验证平台,并比较两个参赛工具与基线。
本文介绍首届 FSE-AIware 工具竞赛,主题为基于智能体的 Python 依赖解析。竞赛为研究者与从业者提供平台,用以构建、测试和验证相关智能体工具;平台包含现成的最先进智能体工具与验证框架,用于评估参赛工具性能。首届竞赛收录来自不同团队的两个工具,并在新生成的、具有不同依赖复杂度的 Python 项目数据集上,与基线工具进行比较。论文概述竞赛背景、框架、参赛工具、评估方法与主要发现,并讨论 Python 依赖解析的性能表现与挑战。
This is the inaugural edition of the FSE-AIware Tool Competition for Agentic Based Python Dependency Resolution. The aim of this competition is to provide a platform for researchers and practitioners to build, test and validate their agentic tools for Python dependency resolution. The platform provides an existing state-of-the-art agentic tool and a validation framework to evaluate the performance of the tools. In this first edition, the competition includes two tools from different teams. All tools were validated on a newly generated dataset of Python projects with various dependency complexity. We compare all tools against the baseline tool and provide insights into the performance of the tools and the challenges of Python Dependency Resolution. This paper provides an overview of the competition, detailing its context, framework, participating tools, evaluation methodology, and key findings.