Skip to content

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

Oct 12 · Monday

All papers

Found 4 papers
Filter5

Attack type

Prompt injection4Poisoning and backdoors4

Defense type

Instruction–data separation3Permissions and approval3

Impact

Unauthorized actions4

System type

AI Agent4

Attack surface / entry point

User input23Web pages and interfaces18Documents and messages19Tool responses17Tool and skill descriptions6Code repositories and dependencies3Retrieval corpus11Memory writes36Inter-agent messages4Images, audio and video20Distribution channels4

Layer

Application layer4Training and data layer4

System component

MCP and skills36Memory3Guardrails and judges5Monitors4Permissions and sandboxing7Fine-tuning and open weights3Embeddings0Multimodal1
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 254papers have no tags yet and are not included in filtered results.
Attack surface / entry point：Distribution channelsLayer：Application layerSystem component：MemorySystem component：EmbeddingsSystem component：MultimodalClear all

Another 254papers do not have category tags yet and are not included in filtered results.

  1. DefensesarXiv 2610.02861

    论文提出在 Kubernetes 上保护 AI Agent 的七层纵深防御框架与参考架构

    提出 Kubernetes 运维智能体的七层纵深防御框架与参考架构

    arXiv
    2610.02861
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Abstract把运维智能体当不可信租户，用七层控制在完全攻破假设下约束行动。

    把 Kubernetes 上的运维智能体当作不可信租户，在假定提示注入已完全控制其工具选择的前提下，用基础设施重建数据面与控制面的边界。要解决的是：智能体读日志、注解、工单或 MCP 工具描述时，这些内容可以变成它带着自身凭据执行的命令。模型对齐和分类器只降低概率。

    Paper analysisRead analysis
  2. Evaluation and benchmarksarXiv 2609.35912

    MMSkillRisk：多模态技能图像中的隐藏指令可诱导 Agent 执行未授权操作

    提出 MMSkillRisk 基准，评测多模态技能图像中的隐藏指令诱导 Agent 越权操作

    arXiv
    2609.35912
    Published
    Layer
    Application layer
    Setting
    coding agent
    Evaluated models
    GPT-5.6-sol, Kimi-K2.6, Kimi-K3 and others6

    Abstract论文提出 MMSkillRisk 基准与 NCVA 攻击，揭示多模态技能中教学图像潜藏指令可引发未授权操作并与良性任务共存。

    Paper analysisRead analysis
  3. DefensesarXiv 2603.12644

    揭示自主 Agent 的安全威胁与防御架构：OpenClaw 案例研究

    梳理自主 Agent 威胁并提出全生命周期防御蓝图 FASA

    arXiv
    2603.12644
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Abstract以 OpenClaw 为例划分三层风险，并用 FASA 描述从输入到系统执行的防御。

    这是一份以 OpenClaw 为案例的威胁梳理和防御架构报告，不是带分数的攻击或防御评测。

    Paper analysisRead analysis
  4. DefensesarXiv 2603.11619

    驯服 OpenClaw：自主 LLM 智能体威胁的安全分析与缓解

    提出对齐五阶段的纵深防御，缓解自主 Agent 跨阶段威胁

    arXiv
    2603.11619
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Abstract这是对 OpenClaw 的生命周期威胁分析与分层缓解设计，案例为定性证据。

    这是对 OpenClaw 的生命周期威胁分析，并配套一套分层缓解设计。要处理的问题是：自主智能体把即时通讯、持久记忆、第三方插件和高权限执行连在一起之后，攻击不再限于单轮越狱，而会跨初始化到执行的多个阶段复合。作者认为只加固单个接口的防御，挡不住这种跨时间攻击。

    Paper analysisRead analysis
You have reached the end