MCP-Persona: Benchmarking LLM Agents on Personalized MCP Tools and Tasks
Zhejiang University · University of Illinois at Urbana-Champaign · University of Oxford · Shanghai Jiaotong University · Shanghai Jiao tong University · Shanghai Jiao Tong University
PDF 由论文原始站点提供，PaperCompass 不保存论文文件。
摘要
a transformative standard for connecting large language models (LLMs) with external data sources and tools, and has been rapidly adopted across personal applications and development platforms. However, existing benchmarks predominantly focus on generic information-seeking tools and fail to capture the practical challenges posed by personal social applications, where tools interact with individual accounts or local databases. To bridge this critical gap, we introduce MCP-Persona, the first benchmark specifically designed for evaluating agent performance on real-world, personalized MCP tools. MCP-Persona encompasses a diverse set of widely-used applications, ranging from social media platforms like Reddit and Xiaohongshu (Rednote) to enterprise collaboration suites such as Lark (Feishu) and Slack. Our extensive experiments on various state-of-the-art (SOTA) agents demonstrate their significant struggles with personalized tool use, thereby highlighting the benchmark's crucial role in identifying and addressing these limitations. MCP-Persona is publicly available at \href{https://anonymous.4open.science/r/MCP-Persona-F85D}{https://anonymous.4open.science/r/MCP-Persona-F85D}