CredLeak-Bench
提出 CredLeak-Bench，评测 Agent 面对钓鱼的凭据泄露与恢复
- arXiv
- 2610.08871
- Published
- Layer
- Application layer
- Setting
- web agent
- Evaluated models
- Claude Sonnet 5, Claude Opus 4.8, Gemini 3.6 Flash and others7
- AttacksPrompt injection
All papers
Filter5
Attack type
Attack technique
Model risks
Impact
System type
Attack surface / entry point
Layer
System component
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 227papers have no tags yet and are not included in filtered results.
Another 227papers do not have category tags yet and are not included in filtered results.
StegoMemory red team testing: agent persistent memory can be used as a cross-session covert channel
提出 StegoMemory，评测智能体持久记忆作为跨会话隐写信道
- arXiv
- 2610.04589
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- Llama 3.1 405B, Claude 3.5 Sonnet, GPT-4o and others13
- AttacksPoisoning and backdoors
- TechniqueEncoding and obfuscation
- ComponentMemory
Abstract研究揭示智能体持久记忆可作为隐蔽信道跨会话传输机密，写入持久化是主要防御瓶颈，呼吁建立记忆完整性防御。
EvoRiskBench: An evolutionary benchmark for runtime security risks of workspace agents
构建 EvoRiskBench 评测工作区 Agent 的运行时安全风险
- arXiv
- 2610.03153
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- GPT-5.6 Sol, DeepSeek-V4-Pro-0813, Claude Opus 5
- AttacksPrompt injection
- ComponentMCP and skills
Abstract论文构建了工作空间智能体运行时风险基准 EvoRiskBench，发现高执行力智能体在 IPI 下面临较高安全风险。
MADBench：多智能体辩论安全性评测基准发布
提出 MADBench，评测多智能体辩论是吸收还是放大对抗影响
- arXiv
- 2609.39146
- Published
- Layer
- Application layer
- Setting
- Multi-agent systems
- Evaluated models
- gpt-4o, GPT-5, GPT-6 Astra
- AttacksPrompt injection
- ComponentRAG
AbstractMADBench 显示辩论能吸收部分准确率攻击，同时放大未授权读写。
MADBench 是一个评测多智能体辩论安全性的基准，用来看攻击是被辩论吸收还是被放大。
pikit：面向间接提示注入研究与评测的可组合工具包
构建可组合工具包 pikit，评测编码智能体的间接提示注入
- arXiv
- 2609.36817
- Published
- Layer
- Application layer
- Setting
- coding agent
- AttacksPrompt injection
Abstractpikit 提供组合式间接注入评测。
pikit 是面向间接提示注入研究的组合式工具包，主要贡献是组织攻击、投递通道、防御和判定流程，而非提出新的攻击优化算法。
CoSec：面向社区场景的 LLM Agent 安全基准
用 CoSec 评测社区 Agent 是否越界披露受保护信息
- arXiv
- 2609.34790
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- OpenClaw with Gemini 3.1 Pro, OpenClaw with Qwen 3.8 Max, OpenClaw with MiniMax M2.5 and others13
- AttacksPrompt injection
- ComponentMemory
研究审计 Agent 安全评测缺陷：工具调用中介在间接提示注入下的有效评测框架
提出卡点评测框架以审计智能体间接提示注入评测的测量缺陷
- arXiv
- 2609.32691
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- gpt-5.6-terra, llama3.1:8b, gemma4:12b and others4
- AttacksPrompt injection
Abstract论文揭示了 IPI 评测中导致指标虚高的系统性缺陷，通过构建有效评测框架纠正了多项安全与模型能力结论。
Cave-Bench：虚假指控下 LLM 智能体最高 60.06% 的运行会破坏正确工作
提出 CAVE-BENCH，评测智能体在虚假指责下破坏已完成的正确工作
- arXiv
- 2609.32616
- Published
- Layer
- Application layer
- Setting
- coding agent
- Evaluated models
- Claude-Sonnet-5, Claude-Opus-5, GPT-5.6-Sol and others14
- AttacksPrompt injection
- RiskDangerous agent actions
Abstract论文提出了评估智能体在虚假指责下破坏已有正确成果的 CAVE-BENCH，发现强模型易推翻已知证据，提出了门控缓解方案。
AgentXploit：面向 AI Agent 的仓库到运行时自动化红队系统
提出 AgentXploit，从代码仓库发现攻击路径并在运行时验证利用
- arXiv
- 2609.31318
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- GPT-4.1
- AttacksPrompt injection
研究提出 ACE 语料库，评估内核级证据对 Agent 安全检测的有效性
构建 ACE 语料，评测内核级证据对 Agent 攻击检测的有效性
- arXiv
- 2609.28915
- Published
- Layer
- Application layer
- Setting
- coding agent
- Evaluated models
- AdaBoost, XGBoost, TabPFN and others10
- AttacksPrompt injection
- ComponentMCP and skills
AbstractACE 用配对的内核与应用层证据表明，syscall 痕迹可判别智能体攻击，两层组合通常更好。
ACE 是一份把智能体会话的内核 syscall 痕迹与应用层清单、转录配对起来的检测语料和基准。。
DUMA-Bench：评估 LLM Agent 安全的双控多智能体基准
提出双控多智能体基准 DUMA-Bench，测量对抗环境下的策略违反
- arXiv
- 2609.24662
- Published
- Layer
- Application layer
- Setting
- Multi-agent systems
- Evaluated models
- GPT-5, GPT-5-mini, GPT-5-nano and others14
- AttacksPrompt injection
AbstractDUMA-Bench 用双控交互评测智能体安全。
DUMA-Bench 把 τ²-bench 的双控交互做成智能体安全评测，用来看策略违反如何随用户、模型和环境一起出现。
研究测量 LLM Agent 记忆投毒防御的良性场景开销
测量记忆投毒防御在良性对话上的效用与 token 开销
- arXiv
- 2609.22818
- Published
- Layer
- Application layer
- Setting
- AI Agent
- AttacksPoisoning and backdoors
- ComponentMemory
Abstract写时代价落在噪声内。
在没有攻击的对话上，给记忆投毒防御计价，而不是再测一轮攻击是否被挡住。
SoK 论文提出 FARSIGHT 框架，评测 15 个学术金融 LLM 交易 Agent 的鲁棒性与安全失败
用 FARSIGHT 方案级评测金融 LLM 交易 Agent 的稳健性与安全失败
- arXiv
- 2609.19705
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- ChatGPT Deep Research
- AttacksPrompt injection
- RiskDangerous agent actions
- ComponentMemory
Emergence World：对长时程多智能体系统开展对抗压力测试
提出 Emergence World，对长时程多智能体系统开展对抗压力测试
- arXiv
- 2609.17320
- Published
- Layer
- Application layer
- Setting
- Multi-agent systems
- Evaluated models
- Claude Opus 4.8, DeepSeek v4 Pro, Gemini 3.5 Flash and others9
- AttacksPrompt injection
- RiskMisalignment and value divergence
- ComponentMemory
Abstract作者通过长期平行世界实验指出模型对齐不具备可组合性，单模型群体会产生社会奉承、协同退缩及语言不透明等失效。
MMPIBench：多模态提示注入对六种 Agent 框架的跨框架评测
用 MMPIBench 评测多模态提示注入对 Agent 框架的影响
- arXiv
- 2609.09404
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- Claude Opus 4.8, GPT-5.4, Gemini 3.1 Pro and others6
- AttacksPrompt injection
- TechniqueCross-modal payloads
- ComponentVision
Abstract论文提出 MMPIBench 评测多模态提示注入对智能体的影响，发现模型主导行为、尝试远超完成，音频通道缺乏防护。
The researchers propose SkillMisevo-Bench to quantify the risk of skill mis-evolution in self-improving agents.
提出 SkillMisevo-Bench，量化自改进 Agent 的技能误演化
- arXiv
- 2608.12851
- Published
- Layer
- Application layer
- Setting
- coding agent
- Evaluated models
- Claude Code, Codex, Hermes and others4
- AttacksPoisoning and backdoors
- TechniqueDormant triggers
- ComponentMemory
ActBench: A Self-Evolving Benchmark for Behavioral Safety of Collaborative Agents
构建自演化基准 ActBench，评测协作智能体的行为安全
- arXiv
- 2608.09476
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- Claude-Opus-4.8, Claude-Sonnet-4.6, GPT-5.5 and others15
- AttacksPrompt injection
- ComponentPermissions and sandboxing
AbstractActBench 通过自演化搜索与双证据重构评测轨迹行为安全，揭示基础模型决策差异主导了协同智能体的操作风险。
多智能体协同过滤系统的连接性攻击与防御研究
复现 AgentCF 上的多智能体攻防并刻画连通性作用
- arXiv
- 2608.03272
- Published
- Layer
- Application layer
- Setting
- Multi-agent systems
- Evaluated models
- Qwen3-235B-A22B, Claude Haiku 4.5, Mixtral 8×7B and others4
- AttacksPoisoning and backdoors
- ComponentMemory
Abstract连通性沿 k 与ρ改变攻防，且用户/物品与早期/稳态需要分开看。
作者在 AgentCF 上研究连通性如何改变多智能体协同过滤的攻防，复现已有方法而不是提出新载荷。
OpenART 提出环境演化红队框架
构建 OpenART 竞技场，以 EMHA 演化持久环境评测智能体安全
- arXiv
- 2608.00677
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- GPT-5.5, Claude-Opus-4.8, GLM-5.2 and others5
- AttacksPrompt injection
- TechniqueAutomated search
MemSecBench: Tracing agent memory poisoning from persistence to consequences and repair
Proposing MemSecBench to evaluate agent memory poisoning and selective repair
- arXiv
- 2607.27080
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- DeepSeek-V4-Pro, MiniMax-M3, GPT-5.5
- AttacksPoisoning and backdoors
- ComponentMemory