Translation in progress Some entries still include text in the original language.
Oct 12 · Monday
All papers Found 18 papers Another 232 papers do not have category tags yet and are not included in filtered results.
“
About this collection Collection in progress We are building a large AI safety paper collection to help researchers find work related to their interests. Papers are collected automatically andfiltered only for relevance to AI safety , without judging quality, novelty or reading value. The default “Annotated first” shows papers with tags or published paper analyses first. Each group is ordered by original publication date, newest first, with undated papers last. “Newest first” uses publication dates only. This order is not a quality ranking.
One-line summaries, tags and paper analyses are prepared by AI from the papers. Language models may have systematic biases and produce hallucinations.We plan to carry out human annotation and review to reduce errors in this content.
提出 PyCache Trap，用字节码缓存替换绕过 Agent 技能扫描器
Published Oct 7
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。
提出 CRIME 框架，组合良性 Agent 技能诱发恶意行为
Published Oct 5
Evaluated models DeepSeek-V4-Flash , GPT-4o-mini , GPT-4o and others6 Abstract 良性技能经 DCA 或 ACA 组合后可产生恶意环境后果，单独审查对不上。CRIME 把公共技能库中各自通过审查的技能配成两两组合，看联合执行能否实现指定恶意任务。 。
提出 PRETEXT 攻击，迭代改写技能文本绕过技能扫描器并让 Agent 执行木马
Published Sep 30
Evaluated models glm (z-ai/glm-5.1) , qwen3t (qwen/qwen3-235b-a22b-thinking-2507) , gpt-oss (openai/gpt-oss-120b) Abstract 论文提出了针对多阶段技能检测器的白盒攻击，揭示了单纯依赖大模型自适应检测的局限。PRETEXT 是一项针对基于“静态规则加语义大模型”的智能体技能安全检测框架的白盒攻击与协同进化评估研究。
提出 AdaLCPI，把恶意指令拆成碎片嵌入工具返回并诱导 Agent 越权执行
Published Sep 29
Evaluated models GPT-4.1 , GPT-5.1 , GPT-5.6-Luna and others7 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
Published Sep 28
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。
提出残余权限重放攻击，跨任务复用长期 Agent 的历史授权绕过确认
Published Sep 28
Evaluated models GPT-4.1 , Gemini-3.1-Flash-Lite , Qwen3-14B and others6 提出 SkillDRE，用预执行与运行时双阶段反馈演化恶意 Agent 技能
Published Sep 26
Evaluated models DeepSeek-V4-Pro , GLM-5.2 , Qwen3.5-397B-A17B and others4 提出条件触发的爆炸提示词，并设计摄入期检测器 DeFuse
Published Sep 19
Evaluated models Llama-3.1-8B-Instruct , Llama-3.1-70B-Instruct , Qwen2.5-32B-Instruct and others15 Abstract 论文揭示了条件触发使提示注入绕过防御并在触发时执行工具，提出轻量检测器并在摄入期有效防范。论文针对大语言模型智能体面临的间接提示注入威胁，研究了利用条件句式实现时间分离的爆炸提示词。
提出 ColluSkill，用跨技能组合绕过 Agent 技能扫描器
Published Aug 10
Evaluated models GPT-5.5 , DeepSeek-V4-Pro , GLM-5.2 Abstract 论文揭示了单技能孤立扫描的盲区，提出多技能链攻击 ColluSkill 与上下文防御 ChainGuard 并验证了有效性。
提出 SYNCHAIN，诱导计算机使用 Agent 自建跨任务持久攻击链
Published Aug 7
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。
提出 Relink 攻击，在 Agent 压缩边界将分散良性片段重组为恶意指令
Published Jun 20
Evaluated models Gemma-4-31B-it , GPT-OSS-20B , Qwen3.6-27B and others15 Abstract 论文指出提示词压缩引入重链漏洞，提出 RELINK 攻击工具与 KBRA 边界审计防御，证明压缩边界应作为关键安全边界。本文研究了智能体基于摘要的提示词压缩所引入的新型安全漏洞——重链（Relinking），并提出了系统化的攻击构建方法与边界审计防御机制。
提出 PoisonedSkills，评测编码 Agent 技能生态的文档投毒
Published Apr 3
Evaluated models Claude Sonnet 4.6 , GLM-4.7 , MiniMax-M2.5 and others5 Abstract 作者称技能文档可成为系统操作攻击入口，但结果只评估加载后的阶段。PoisonedSkills 研究编码智能体的技能供应链投毒，重点是文档内容如何进入实际系统操作。 。
Propose Blindfold, jailbreaking embodied LLMs with action-level manipulation
Published Mar 2
Evaluated models Llama-3.1-8B , DeepSeek-R1-14B , Gemma-3-27B and others8 Abstract Blindfold jailbreaks embodied LLMs with action-level proxy planning, and the authors claim it can bypass semantic checks and is effective in both simulation and real-world settings.Blindfold is a no-box attack against embodied LLM planning modules, rewriting malicious intent into seemingly benign action sequences.
Proposes SkillJect, which automatically generates poisoned skills to induce code agents to execute hidden scripts
Published Feb 16
Evaluated models Claude-4.6-Sonnet , GPT-5-mini , GLM-4.7 and others10 Abstract SkillJect poisons skills using dual channels and closed-loop refinement.SkillJect is an automated prompt injection framework targeting installable agent skills. The authors address the problem that malicious instructions placed in publicly available or manually installed skills are often rejected or ignored, yet once a skill is installed, it is repeatedly loaded as trusted guidance.
Attacks arXiv 2510.04885 · Oct 6, 2025 提出 RL-Hammer，从零训练攻击模型经工具输出实施提示注入
Published Oct 6, 2025
Evaluated models Llama-3.1-8B-Instruct , Meta-SecAlign-8B , Meta-SecAlign-70B and others10 Abstract 作者称通过无热启动的强化学习可有效突破具备防御的商业模型，论文提出的 RL-Hammer 促使业界关注提示注入风险。RL-Hammer 是一个无需热启动数据、基于 GRPO 算法从零训练的黑盒提示注入对抗攻击框架。现有针对大语言模型智能体的提示注入防御（如 Instruction Hierarchy 与 SecAlign）主要在静态或弱攻击下测试，在面对自动化强力攻击时的真实防御能力缺乏充分检验。方法的核心机制包括。
Attacks arXiv 2509.25624 · Sep 30, 2025 提出 STAC 框架，把恶意目标拆成多轮看似无害的工具调用链
Published Sep 30, 2025
Evaluated models GPT-4.1-2025-04-14 , GPT-4.1-mini-2025-04-14 , Qwen3-32B and others8 Abstract 作者提出并验证了智能体在序列化工具攻击下的普遍脆弱性，指出针对累积动作序列防御的必要性。
Attacks arXiv 2509.22040 · Sep 26, 2025 通过规则文件提示注入劫持编码编辑器并执行未授权终端命令
Published Sep 26, 2025
Evaluated models Cursor , GitHub Copilot , Claude 4 Sonnet and others4 Abstract 论文提出了针对代码编辑器的提示注入评估框架 AIShellJack，实验显示攻击能诱导 AI 执行未授权终端命令。
Attacks arXiv 2507.10457 · Jul 15, 2025 提出 LPCI，把延迟条件载荷写入记忆与检索库以跨会话触发
Published Jul 15, 2025
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。
You have reached the end