{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2606.02470", "kind": "paper", "url": "https://arxiv.org/abs/2606.02470"}, {"id": "gh:wwh0411/mcp-persona", "kind": "repo", "url": "https://github.com/wwh0411/MCP-Persona"}], "authors": ["Wenhao Wang", "Peizhi Niu", "Gongyi Zou", "Xiyuan Yang", "Jingxing Wang", "Haoting Shi", "Yaxin Du", "Jingyi Chai", "Xianghe Pang", "Shuo Tang", "Yanfeng Wang", "Siheng Chen"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2606.02470", "source_url": "https://arxiv.org/abs/2606.02470"}, "categories": ["Agents", "Agents & Tool Use", "General AI"], "description": {"en": "MCP-Persona evaluates LLM agents on real-world personalized MCP tools across social media, collaboration, email, and content management applications. It includes 173 tool-chain tasks, 139 unique tools, and 18 MCP servers, with a fully automated environment simulation pipeline for reproducible evaluation."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2606.02470", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2606.02470", "title": "MCP-Persona"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2606.02470", "modality": null, "name": "MCP-Persona", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"confidence": "High", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-10-14T09:44:51Z", "data_status": "primary-source-indexed", "display_eligible": "unknown", "origin_record_id": "bm_mcp-persona_e4dd478c", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "39942567d0de8600d59063b55c264f97b77c7d66f59b68b1849a2d524542cadb", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2606.02470"}, "publisher": null, "released": "2026-06-01", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2606.02470", "source_url": "https://arxiv.org/abs/2606.02470"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2606-02470", "source": "claire_radar", "source_benchmark_id": "2606.02470", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Agents & Tool Use", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": "https://github.com/wwh0411/MCP-Persona", "githubScope": "benchmark_repo", "githubStars": 8, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 14, "hfPaperUrl": "https://huggingface.co/papers/2606.02470", "observedAt": "2026-10-04T15:31:48.030916Z", "signalStatus": {"githubStars": {"observedAt": "2026-10-04T15:31:48.030916Z", "state": "fresh"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-10-04T15:31:48.030916Z", "lastSuccessfulDate": "2026-09-30", "state": "stale"}}, "status": "fresh"}, "capabilities": [], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "4521a2d9c11d1d10cc5db2c38447b4f940b0a15a163d1df64fa97ed9d8ab8741", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "MCP-Persona evaluates LLM agents on real-world personalized MCP tools across social media, collaboration, email, and content management applications. It includes 173 tool-chain tasks, 139 unique tools, and 18 MCP servers, with a fully automated environment simulation pipeline for reproducible evaluation.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["exact coined title identity tied to benchmark evidence", "benchmark release stated in one sentence", "public artifact URL"], "snippet": "To bridge this critical gap, we introduce MCP-Persona, the first benchmark specifically designed for evaluating agent performance on real-world, personalized MCP tools."}, "familyId": "bmf_d0faf0a6b15e", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [{"evidence": [{"basis": "existing-code-link", "linkedUrl": "https://github.com/wwh0411/MCP-Persona", "url": "https://github.com/wwh0411/MCP-Persona"}, {"basis": "existing-metric-association", "linkedUrl": "https://github.com/wwh0411/MCP-Persona", "url": "https://huggingface.co/papers/2606.02470"}, {"basis": "linked-from-report", "linkedUrl": "https://github.com/wwh0411/MCP-Persona", "url": "https://arxiv.org/abs/2606.02470"}, {"basis": "linked-from-report", "linkedUrl": "https://github.com/wwh0411/MCP-Persona.", "url": "https://arxiv.org/abs/2606.02470"}], "scope": "benchmark_repo", "url": "https://github.com/wwh0411/MCP-Persona", "verification": "source-linked"}], "status": "linked"}, "heat": null, "id": "bm_mcp-persona_e4dd478c", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": "https://github.com/wwh0411/MCP-Persona", "data": null, "hfPaper": "https://huggingface.co/papers/2606.02470", "paper": "https://arxiv.org/abs/2606.02470", "pdf": "https://arxiv.org/pdf/2606.02470", "project": null, "report": "https://arxiv.org/abs/2606.02470"}, "metrics": [], "motivation": "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.", "name": "MCP-Persona", "oneLine": "MCP-Persona evaluates LLM agents on real-world personalized MCP tools across social media, collaboration, email, and content management applications. It includes 173 tool-chain tasks, 139 unique tools, and 18 MCP servers, with a fully automated environment simulation pipeline for reproducible evaluation.", "paperTitle": "MCP-Persona: Benchmarking LLM Agents on Real-World Personal Applications via Environment Simulation", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2606.02470", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "publishers": [{"name": "MCP-Persona team", "organizationType": "academic-lab", "role": "benchmark-publisher", "sourceUrl": "https://github.com/wwh0411/MCP-Persona"}], "ranking": {}, "readiness": "Runnable", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-06-01", "source": {"authors": ["Wenhao Wang", "Peizhi Niu", "Gongyi Zou", "Xiyuan Yang", "Jingxing Wang", "Haoting Shi", "Yaxin Du", "Jingyi Chai", "Xianghe Pang", "Shuo Tang", "Yanfeng Wang", "Siheng Chen"], "categories": ["cs.AI"], "id": "2606.02470", "title": "MCP-Persona: Benchmarking LLM Agents on Real-World Personal Applications via Environment Simulation", "type": "arxiv", "url": "https://arxiv.org/abs/2606.02470"}, "sourceUpdatedAt": "2026-06-01T16:44:10Z", "topics": ["Agents"], "venueAttempts": [], "whyItMatters": "Existing benchmarks overlook personalized MCP tool use, which is critical for practical agent deployment. MCP-Persona provides a sandboxed, reproducible environment to test agents on realistic personal tasks, helping identify limitations and guiding development of more capable agents."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
