PulseAugur
中
实时 20:12:59
English(EN) ShadowBench-MCP: Measuring Shadow Invocations in MCP-Style Tool Routing

新基准衡量LLM代理在工具路由中的安全风险

研究人员开发了ShadowBench-MCP,这是一个旨在衡量大型语言模型(LLM)代理中“影子调用”的新基准,这些代理充当工具和API的路由器。当LLM选择更危险的执行路径,可能绕过安全控制时,即使请求并非恶意,也会发生这种现象。该基准将任务保留、操作压力和路由几何漂移分开,以计算执行前的SODS分数。在Groq上托管的qwen/qwen3-32b路由器上进行的早期测试表明,效率优先路由导致了显著的影子调用率,而策略感知路由则大大降低了这一比率。 AI

影响 引入了一种评估和潜在改进充当工具路由器的LLM代理安全性的新方法。

排序理由 该集群包含一篇学术论文,介绍了用于评估LLM代理安全性的新基准和方法论。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.MA (Multiagent) 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

新基准衡量LLM代理在工具路由中的安全风险

本文如何被排名

Signal score
1 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
该集群包含一篇学术论文,介绍了用于评估LLM代理安全性的新基准和方法论。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
paper, safety
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
1 days old
Coverage has settled into its steady-state source set.

完整方法见我们的编辑标准。

报道来源 [1]

  1. arXiv cs.MA (Multiagent) TIER_1 English(EN) · Wenjun Yang ·

    ShadowBench-MCP: 测量 MCP 风格工具路由中的影子调用

    LLM agents increasingly act as routers for tools, APIs, and software-supply-chain workflows, making an execution path - rather than generated text alone - a security-relevant model output. We study shadow invocation: a task-preserving request variant that causes an LLM router to …