arXiv cs.AI· Yitian Zheng, Zhaorun Chen, Haoyu Li, Bo Li·· 12 小时前AI 评分44
GatewayBench：面向 MCP 网关的统一能力与安全评测平台
GatewayBench: Unified Capability and Security Evaluation Platform for MCP Gateways
AI 导读
研究者推出 GatewayBench，首个针对 MCP 网关的统一评测框架与基准，将网关架构划分为扁平式与分层式两类。该基准包含六大领域 524 个任务，工具池来自 251 个真实 MCP 服务器的 11,118 个工具，并注入任务无关工具以模拟大规模部署。
正文 · 原文
Abstract:AI agents have increasingly interacted with external environments by invoking tools, with Model Context Protocol (MCP) emerging as a standard interface for connecting agents to heterogeneous tools and services. As organizations deploy agents at scale, MCP gateways have become a critical middleware layer for aggregating, governing, and shaping the tool surface exposed to agents. Despite the growing importance of MCP gateways in agentic systems, no prior work has systematically studied how gateway design affects agent capability, efficiency, and security. To address this gap, we introduce GatewayBench, the first unified evaluation framework and benchmark for MCP gateways. We characterize gateway architectures into two major paradigms: flattened gateways and hierarchical gateways. GatewayBench comprises 524 tasks across six domains, together with a pool of 11,118 real-world tools from 251 authentic MCP servers. To emulate large-scale deployments, we systematically inject task-irrelevant tools sampled from this pool into each agent environment, enabling controlled evaluation as the accessible tool surface scales. Based on GatewayBench, we conduct extensive experiments across seven widely deployed MCP gateways and diverse agent configurations, covering both benign and adversarial tasks. Our results reveal substantial differences across gateway designs in task completion, tool-use efficiency, and resilience to attacks, and further demonstrate that these effects interact strongly with the underlying model and agent harness. These findings highlight the MCP gateway as a consequential yet underexplored component of the agentic AI stack. By enabling controlled evaluation of gateway architectures under realistic tool ecosystems, GatewayBench provides both practical guidance for MCP gateway design and a rigorous testbed for developing more capable, efficient, and secure agentic systems.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.23449 [cs.AI] |
| (or arXiv:2610.23449v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.23449 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Yitian Zheng [view email]
[v1]
Fri, 16 Oct 2026 13:31:09 UTC (1,428 KB)
来源：arXiv cs.AI · arxiv.org