ICML 2026</p>\n","updatedAt":"2026-06-02T05:46:29.231Z","author":{"_id":"64faa838da4429f25dd19471","avatarUrl":"/avatars/635f33920a05f63ba7783c3ecc75a462.svg","fullname":"WENHAO WANG","name":"wwh0411","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":2,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.6255497336387634},"editors":["wwh0411"],"editorAvatarUrls":["/avatars/635f33920a05f63ba7783c3ecc75a462.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2606.02470","authors":[{"_id":"6a1e585b808ddbc3c7d43dd7","name":"Wenhao Wang","hidden":false},{"_id":"6a1e585b808ddbc3c7d43dd8","name":"Peizhi Niu","hidden":false},{"_id":"6a1e585b808ddbc3c7d43dd9","name":"Gongyi Zou","hidden":false},{"_id":"6a1e585b808ddbc3c7d43dda","name":"Xiyuan Yang","hidden":false},{"_id":"6a1e585b808ddbc3c7d43ddb","name":"Jingxing Wang","hidden":false},{"_id":"6a1e585b808ddbc3c7d43ddc","name":"Haoting Shi","hidden":false},{"_id":"6a1e585b808ddbc3c7d43ddd","name":"Yaxin Du","hidden":false},{"_id":"6a1e585b808ddbc3c7d43dde","name":"Jingyi Chai","hidden":false},{"_id":"6a1e585b808ddbc3c7d43ddf","name":"Xianghe Pang","hidden":false},{"_id":"6a1e585b808ddbc3c7d43de0","name":"Shuo Tang","hidden":false},{"_id":"6a1e585b808ddbc3c7d43de1","name":"Yanfeng Wang","hidden":false},{"_id":"6a1e585b808ddbc3c7d43de2","name":"Siheng Chen","hidden":false}],"publishedAt":"2026-06-01T00:00:00.000Z","submittedOnDailyAt":"2026-06-02T00:00:00.000Z","title":"MCP-Persona: Benchmarking LLM Agents on Real-World Personal Applications via Environment Simulation","submittedOnDailyBy":{"_id":"64faa838da4429f25dd19471","avatarUrl":"/avatars/635f33920a05f63ba7783c3ecc75a462.svg","isPro":false,"fullname":"WENHAO WANG","user":"wwh0411","type":"user","name":"wwh0411"},"summary":"The Model Context Protocol (MCP) has emerged as 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 https://github.com/wwh0411/MCP-Persona}{https://github.com/wwh0411/MCP-Persona.","upvotes":10,"discussionId":"6a1e585b808ddbc3c7d43de3","githubRepo":"https://github.com/wwh0411/MCP-Persona","githubRepoAddedBy":"user","ai_summary":"MCP-Persona benchmark evaluates agent performance on personalized tools interacting with individual accounts and local databases, revealing significant challenges in current SOTA agents.","ai_keywords":["Model Context Protocol","large language models","external data sources","tools","personalized tool use","agent performance","benchmark","social media platforms","enterprise collaboration suites"],"githubStars":4},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"64faa838da4429f25dd19471","avatarUrl":"/avatars/635f33920a05f63ba7783c3ecc75a462.svg","isPro":false,"fullname":"WENHAO WANG","user":"wwh0411","type":"user"},{"_id":"680b79808cd5360594d733f0","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/62HhHpYdgNVEH_cYBVGOj.png","isPro":false,"fullname":"Xiyuan Yang","user":"XiyuanYang","type":"user"},{"_id":"6857995b242a5555be45c781","avatarUrl":"/avatars/d33c99cd9325b69054e7ca37f194931f.svg","isPro":false,"fullname":"Hamilton-Sht","user":"tingtingg","type":"user"},{"_id":"68af7a626ed35e7370f24ecd","avatarUrl":"/avatars/b28b07b18b771c1f3fe983b8dd7b6989.svg","isPro":false,"fullname":"Peizhi Niu","user":"npzliekon","type":"user"},{"_id":"69cdfefef77cbcf0280997ea","avatarUrl":"/avatars/ed366f6a3f62f18e5a73aa5fe6062dce.svg","isPro":false,"fullname":"Yujie Ge","user":"hereiswendy","type":"user"},{"_id":"65352517fca2c10e43035e32","avatarUrl":"/avatars/f404cef8a79f27435865ac4d85ef0927.svg","isPro":false,"fullname":"xinyuzhu","user":"xinyuzhu","type":"user"},{"_id":"63ecd42a60ff4b318ad1ef47","avatarUrl":"/avatars/c05001fa9c41f713d1fe31d11212faae.svg","isPro":false,"fullname":"RuiYe","user":"ruiye-sjtu","type":"user"},{"_id":"63ca8e060609f1def7e6548a","avatarUrl":"/avatars/1da7947840cb87d5f77c0af9ee11f9c2.svg","isPro":true,"fullname":"Yi Jung","user":"YJ-142150","type":"user"},{"_id":"697e3b1346330c9edc62fd54","avatarUrl":"/avatars/53e092676d569a8eeec5f4496e9a12df.svg","isPro":false,"fullname":"ZOU GONGYI","user":"GlyphZ","type":"user"},{"_id":"69bcfa77dfac620f179229b4","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/noauth/55YI18MOp6RlzL6OQ-e0F.jpeg","isPro":false,"fullname":"萧诗雨","user":"liangziyu1","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2606/2606.02470.md"}">

Get this paper in your agent:

Don't have the latest CLI?

Cite arxiv.org/abs/2606.02470 in a model README.md to link it from this page.

Cite arxiv.org/abs/2606.02470 in a dataset README.md to link it from this page.

Cite arxiv.org/abs/2606.02470 in a Space README.md to link it from this page.

Add this paper to a collection to link it from this page.