Skip to content

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

Oct 12 · Monday

All papers

Found 11 papers
Filter2

Attack type

Jailbreaks50Prompt injection33Poisoning and backdoors11Detection evasion22Malicious use3

Attack technique

Defense-targeted5Dormant triggers3

Defense type

Detection and filtering4

Impact

Unauthorized actions5Output manipulation4

System type

AI Agent8

Attack surface / entry point

User input3Tool responses3Tool and skill descriptions5Distribution channels5

Layer

Application layer8Training and data layer3

System component

MCP and skills40Memory37RAG19Guardrails and judges11Monitors7Reasoning chains4Permissions and sandboxing11Fine-tuning and open weights28Vision8Image generation3Embeddings3Multimodal14
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 305papers have no tags yet and are not included in filtered results.
Attack type：Poisoning and backdoorsSystem component：Guardrails and judgesClear all

Another 305papers do not have category tags yet and are not included in filtered results.

  1. AttacksarXiv 2610.10612

    PyCache Trap 利用 Python 缓存替换绕过七款 Agent Skill 扫描器

    提出 PyCache Trap，用字节码缓存替换绕过 Agent 技能扫描器

    arXiv
    2610.10612
    Published
    Layer
    Application layer
    Setting
    coding agent
    Evaluated models
    CISCO, HSS, AUDITOR and others8
    Abstract论文剖析智能体技能审查与执行脱节盲区，提出 PyCache Trap 攻击并构建基于可信复现的 EAV 防御。

    这篇论文研究了智能体技能（Agent Skills）安全扫描中的“审查–执行脱节”问题，并提出了对应的攻击验证与防御方案。

    Paper analysisRead analysis
  2. AttacksarXiv 2610.07723

    研究者提出答案侧后门：让模型自生成触发词绕过安全拒答

    提出回答端后门，让模型自生成触发词以在干净有害请求下绕过拒答

    arXiv
    2610.07723
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Evaluated models
    Llama-2-7B-chat-hf, Mistral-7B-Instruct-v0.3, Gemma-7B-it and others4

    Abstract论文提出了多轮对话中回答端自生成触发词绕过安全拒绝的后门攻击，指出输入端防护存在盲区。

    Paper analysisRead analysis
  3. DefensesarXiv 2610.07657

    DEFER1 多智能体防御研究报告受控测试中的攻击成功率下降

    实现 DEFER，以确定性检查优先拦截多智能体流水线的内容攻击

    arXiv
    2610.07657
    Published
    Layer
    Application layer
    Setting
    Multi-agent systems
    Evaluated models
    Qwen3-235B-A22B-Instruct-2507, gpt-oss-120b, Llama-4-Scout-17B-16E-Instruct and others5
    AbstractDEFER 把确定性检查放在 judge 之前。

    DEFER 用来测量多智能体授权流水线里，确定性规则在何处结束、LLM judge 从何处开始。

    Paper analysisRead analysis
  5. AttacksarXiv 2609.30383

    研究者提出技能级联攻击

    提出技能级联攻击，把有害目标拆散到多个技能里绕过单技能扫描

    arXiv
    2609.30383
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    GPT-5.4, GPT-5.4 Mini, Claude Opus 4.6 and others11

    Abstract论文探讨了多技能智能体系统面临的技能级联攻击风险，提出自动化红队框架与基准集，指出跨技能交互存在安全防护盲区。

    Paper analysisRead analysis
  9. Evaluation and benchmarksarXiv 2606.15899

    SkillVetBench：用 LLM-as-Judge 评估开源 LLM Agent 技能的多维安全风险

    提出 SkillVetBench 评估开源 Agent 技能的多维安全风险

    arXiv
    2606.15899
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    Qwen2.5-14B-Instruct, Qwen2.5-32B, Llama-3.1-70B and others6

    Abstract公开排行榜用 LLM-as-Judge 给技能分列 SARS 与 CVSS，并对照市场官方审查。

    Paper analysisRead analysis
  10. AttacksarXiv 2604.03890

    从提示到物理动作：针对 LLM 中介机器人控制系统的结构化后门攻击

    提出结构化 JSON 后门，经投毒 LoRA 在机器人控制中触发恶意动作

    arXiv
    2604.03890
    Published
    Layer
    Training and data layer
    Setting
    Embodied systems and robotics, LLM applications
    Evaluated models
    Llama-3.1-8B-Instruct, Gemma-2-9B-IT, DeepSeek-R1-Distill-Llama-8B and others4

    Abstract推理层后门传不到执行。

    Paper analysisRead analysis
  11. AttacksarXiv 2602.01658

    高维离线 Bandit 的高效对抗攻击

    提出三类攻击，训练前扰动奖励模型以劫持离线 bandit

    arXiv
    2602.01658
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Evaluated models
    Image Reward, Aesthetic Model
    Abstract公开奖励权重可被小扰动劫持离线 bandit，打乱日志只部分缓解。

    作者研究离线 bandit 在奖励模型被事先改权重时，还会不会选出最优臂。

    Paper analysisRead analysis
You have reached the end