arXiv cs.AI· Xuanang Gao, Kepeng Lin, Xinyan Shi, Yong Deng, Shuheng Shen, Changhua Meng, Xiaofeng Wu, Jie Yang, Wei Liu·· 1 天前AI 评分34
EvoMCP：让 MCP 智能体与训练数据协同进化
EvoMCP: Co-evolving MCP agents and training data
AI 导读
EvoMCP 提出一种智能体与数据互相进化的框架，通过任务进化、轨迹进化和监督微调的闭环，让 MCP 智能体持续针对自身能力缺口生成训练数据。在 Toolathlon、MCPMark 和 MCP-Atlas 上，该方法相比初始学生模型分别取得 16.5%、10.2% 和 12.6% 的绝对提升。
正文 · 原文
Abstract:Supervised fine-tuning of Model Context Protocol (MCP) agents is typically conducted on a static data distribution, which fails to adapt to the student's evolving capabilities and leaves newly exposed weaknesses insufficiently addressed. To address these challenges, we propose \method, an agent--data mutual evolution framework that enables MCP agents to improve through a closed loop of task evolution, trajectory evolution, and supervised fine-tuning. Specifically, \method\ assesses the student's established capabilities and capability gaps from rollout feedback to identify learning targets. Task evolution then translates these targets into executable tasks by adapting the instructions and observable environment states of compatible verified seed tasks while keeping their verifiers aligned. To provide reliable and efficient demonstrations for these tasks, trajectory evolution uses reusable skills to recover failed teacher rollouts and refine successful ones. The resulting task--trajectory pairs are used to fine-tune the student, while accepted tasks seed subsequent evolution. To close the loop, we re-evaluate the updated student and use its feedback to guide the next round of data evolution. Experiments on Toolathlon, MCPMark, and MCP-Atlas demonstrate consistent improvements over the initial student, yielding absolute gains of 16.5\%, 10.2\%, and 12.6\%, respectively. Project page: this https URL.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.41479 [cs.AI] |
| (or arXiv:2610.41479v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.41479 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Kepeng Lin [view email]
[v1]
Fri, 30 Oct 2026 09:35:29 UTC (946 KB)
来源：arXiv cs.AI · arxiv.org