Skip to content

Translation in progress Some entries still include text in the original language.

Oct 12 · Monday

All papers

Found 2 papers
Filter3

Model risks

Dangerous agent actions1Dangerous capabilities1

System component

MCP and skills2Memory6Guardrails and judges7Monitors14Permissions and sandboxing25Vision3Multimodal4
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 265papers have no tags yet and are not included in filtered results.
Model risks：Dangerous agent actionsModel risks：Dangerous capabilitiesSystem component：MCP and skillsClear all

Another 265papers do not have category tags yet and are not included in filtered results.

  1. OtherarXiv 2609.18120

    PentestChain：面向免费层 LLM 的成本感知 MCP 编排自动化渗透测试框架

    提出成本感知的 MCP 编排框架，用确定性利用图与免费层模型做自动化渗透测试

    arXiv
    2609.18120
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Abstract确定性利用图加免费层级联在十次自建运行上测得付费成本为零。

    PentestChain 是一个十阶段、经 MCP 暴露的自动渗透框架，用确定性利用图把 7B 本地模型挡在利用关键路径之外，并把每次交战的美元成本当作实测指标。。

    Paper analysisRead analysis
  2. DefensesarXiv 2609.12001

    Pheo 提出 OATS：用确定性解析器在运行时治理 Agent 技能动作

    提出 OATS，在工具调用前用确定性解析器治理 Agent 技能动作

    arXiv
    2609.12001
    Published
    Layer
    Application layer
    Setting
    AI Agent

    Abstract论文论证了发布时恶意扫描无法替代运行时动作许可控制，提出了轻量确定性控制门 OATS，展示了双层防御组合的必要性。

    Paper analysisRead analysis
You have reached the end