Skip to content

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

Oct 12 · Monday

All papers

Found 9 papers
Filter3

Attack type

Prompt injection5Poisoning and backdoors9

Attack technique

Automated search18Encoding and obfuscation7Cross-modal payloads1Defense-targeted8Dormant triggers8Selection manipulation10

Impact

Unauthorized actions6Data and privacy leaks3

System type

AI Agent9coding agent5

Attack surface / entry point

Tool and skill descriptions6Memory writes3Distribution channels6

Layer

Application layer9

System component

MCP and skills9Memory22RAG6Guardrails and judges14Reasoning chains4Permissions and sandboxing3Fine-tuning and open weights18Vision49Audio9Video3Image generation6Multimodal64
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.
Attack technique：Cross-modal payloadsAttack technique：Dormant triggersSystem component：MCP and skillsClear all

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

  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
  5. AttacksarXiv 2608.25776

    EvoMal 攻击利用自我进化编码 Agent 的自我投毒实现蠕虫式传播

    提出 EVOMAL，诱导自演化编码 Agent 在创建技能时自我投毒并蠕虫传播

    arXiv
    2608.25776
    Published
    Layer
    Application layer
    Setting
    coding agent
    Evaluated models
    Devstral-Small-2, Gemma-4-31B-IT, Qwen3-Coder-Next and others7

    Abstract论文指出了自演化智能体模仿检索代码的自我毒化漏洞，提出 EVOMAL 攻击与防御，阐明了供应链新风险。

    Paper analysisRead analysis
  6. AttacksarXiv 2608.09577

    ElasticBack：通过触发词与规则耦合在 LLM Agent 技能中植入条件后门

    提出 ElasticBack，在单个 Agent 技能中植入条件后门

    arXiv
    2608.09577
    Published
    Layer
    Application layer
    Setting
    coding agent
    Evaluated models
    GLM-5.2, MiniMax-M3, GPT-5.4 and others4
    AbstractElasticBack 用冻结规则加遗传搜索演化 trigger，在单 skill 上做条件后门。

    ElasticBack 是放在 LLM agent 单个 skill 里的条件后门攻击。。

    Paper analysisRead analysis
  7. AttacksarXiv 2608.06862

    SynChain：诱导计算机使用 Agent 自建攻击链并跨任务持久化

    提出 SYNCHAIN，诱导计算机使用 Agent 自建跨任务持久攻击链

    arXiv
    2608.06862
    Published
    Layer
    Application layer
    Setting
    Browser and computer use, coding agent
    Evaluated models
    Qwen3.5-9B, Llama-3.1-8B, Ministral-3-8B

    Abstract论文揭示了 CUA 自合成工件带来的内部供应链威胁，验证了潜伏载荷的跨任务传播能力与防御局限。

    Paper analysisRead analysis
  8. AttacksarXiv 2608.05563

    研究者提出 PoisonedEvolution 轨迹投毒攻击

    提出 PoisonedEvolution 轨迹投毒，诱导自演化技能系统固化恶意行为

    arXiv
    2608.05563
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    GPT-5.4, MiniMax-M2.5, DeepSeek-V3.2 and others6
    Abstract论文揭示自演化技能系统的证据晋升漏洞，提出并验证轨迹投毒攻击。

    这篇论文揭示了自演化技能系统在将交互经验转化为持久指令时的安全风险。

    Paper analysisRead analysis
  9. AttacksarXiv 2507.10457

    论文提出 LPCI：针对 Agent 逻辑层与持久记忆的提示控制注入攻击

    提出 LPCI，把延迟条件载荷写入记忆与检索库以跨会话触发

    arXiv
    2507.10457
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    ChatGPT, Claude, LLaMA3 and others8
    Abstract作者定义 LPCI。

    作者把LPCI定义为针对智能体逻辑执行、记忆留存和工具集成的跨阶段漏洞类，而不是单次输入上的提示改写。

    Paper analysisRead analysis
You have reached the end