DefensesarXiv 2610.09935
AgentTracer：通过细粒度意图-执行对齐追踪间接提示注入攻击
提出 AgentTracer，事后定位间接提示注入的注入点并重建攻击链
- arXiv
- 2610.09935
- Published
- Layer
- Application layer
- Setting
- AI Agent
- DefensesMonitoring and auditing
- AttacksPrompt injection
All papers
Filter4
Attack type
Model risks
Defense type
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 232papers have no tags yet and are not included in filtered results.
Another 232papers do not have category tags yet and are not included in filtered results.
- DefensesarXiv 2610.09793
用时序逻辑监控工具 Agent
提出 MFOTL 离线时序验证以监控工具 Agent 的危险动作链
- arXiv
- 2610.09793
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Attacker
- Black-box
- DefensesMonitoring and auditing
- AttacksPrompt injection
- ComponentMonitors
Abstract形式化验证可有效识别多步攻击，但其有效性高度依赖轨迹的可信历史记录。
这篇论文对工具调用大语言模型智能体的安全监控进行了形式化时序验证研究。
研究者提出统一滥用监控基准，覆盖分解攻击与提示注入约 6200 条轨迹
提出危害窗口与区间指标，统一评测分解攻击和提示注入下的智能体滥用监视
- arXiv
- 2610.07089
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- Gemma 4 26B-A4B, Qwen3.6 27B, Qwen3.5 27B and others12
- DefensesMonitoring and auditing
- AttacksJailbreaks
- ComponentMonitors
Abstract统一智能体滥用监视框架，动作视角兼顾两类威胁但危害定位仍存瓶颈。
- DefensesarXiv 2610.05163
Researchers construct segmented prompt injection on Claude Code and Codex and release a boundary action audit benchmark
提出 PAA 方法，在边界动作前审计长流程智能体的分阶段提示注入
- arXiv
- 2610.05163
- Published
- Layer
- Application layer
- Setting
- coding agent
- Evaluated models
- Claude Code (v2.1.172, Claude Sonnet 5), Codex CLI (v0.144.0, GPT-5.5)
- DefensesMonitoring and auditing
- AttacksPrompt injection
- ComponentMonitors
Abstract论文针对长流程注入提出边界动作审计与 PAA，在保持低误阻的同时实现高拦截率。
本研究针对长流程智能体在执行任务时面临的分阶段间接提示注入威胁，提出了在动作生效前进行拦截的边界动作审计框架与路径对齐归因方法。
- DefensesarXiv 2610.00400
DART: Detecting emerging safety risks in multi-turn LLM agents with representation shift
提出 DART，用去噪表征转移检测多轮智能体攻击并注入提醒
- arXiv
- 2610.00400
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Attacker
- Black-box
- Evaluated models
- Qwen3-8B, Qwen3-14B, Qwen3-32B and others6
- DefensesMonitoring and auditing
- AttacksJailbreaks
- ComponentMonitors
- DefensesarXiv 2609.39548
NDDL：用正常扩散动力学为文生图模型做后门防御
提出 NDDL，用良性扩散轨迹检测文生图后门提示并定位触发 token
- arXiv
- 2609.39548
- Published
- Layer
- Model layer
- Setting
- Models and APIs
- Evaluated models
- Stable Diffusion v1.5, Stable Diffusion XL, Stable Diffusion v3.5 and others4
- DefensesMonitoring and auditing
- AttacksPoisoning and backdoors
- ComponentImage generation
AbstractNDDL 从良性扩散轨迹学习正常转移，用动态不一致检测后门并定位触发 token。
NDDL 是面向文生图扩散模型的后门防御：防御者只有白盒模型访问和良性提示词，没有后门先验，需要区分后门提示词并找出触发 token。
- DefensesarXiv 2609.35659
Tracekit：面向自主编码 Agent 的防篡改意图-推理-动作审计系统
提出 Tracekit，审计编码 Agent 的意图、推理与动作并检测账本篡改
- arXiv
- 2609.35659
- Published
- Layer
- Application layer
- Setting
- coding agent
- Evaluated models
- Claude Code (claude -p, version 2.1), 独立评审（同一 CLI，无工具）
- DefensesMonitoring and auditing
- AttacksPrompt injection
- RiskDangerous agent actions
- ComponentMonitors
AbstractTracekit 把意图、自述和动作写入可锚定的哈希链。
Tracekit 是无第三方依赖的编码智能体审计层：把人类请求、模型自述和已执行动作并排写入可外部锚定的哈希链，执行前用正则门，事后用规则和独立模型互查。
- DefensesarXiv 2609.14987
ActGuard：面向 LLM Agent 间接提示注入的执行前动作审计框架
提出 ActGuard 执行前动作审计，定位并遮盖间接提示注入的恶意片段
- arXiv
- 2609.14987
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- GPT-4o-mini, Gemini-2.5-Flash, DeepSeek-V4-Flash
- DefensesMonitoring and auditing
- AttacksPrompt injection
- DefensesarXiv 2608.05578
AMS 工具通过激活分析检测语言模型的安全训练改动
提出 AMS，用激活分离与方向相似度检测安全训练改动
- arXiv
- 2608.05578
- Published
- Layer
- Model layer
- Setting
- Models and APIs
- Evaluated models
- Llama-3.1-8B-Instruct, Llama-3.2-3B-Instruct, Llama-3.2-1B-Instruct and others14
- DefensesMonitoring and auditing
- AttacksModel tampering
- ComponentFine-tuning and open weights
AbstractAMS 按激活签名把安全改动分四类，前三类可由两级扫描覆盖，行为微调会漏。
AMS 是面向开放权重模型、部署前使用的激活扫描器，用来看安全训练被拿掉或改写之后，激活空间里还留着什么几何痕迹。
- DefensesarXiv 2608.02683
S³：用多阶段防御提升 LLM 智能体安全
提出 S³，用 guard agent 编排阶段安全技能防御智能体多阶段风险
- arXiv
- 2608.02683
- Published
- Layer
- Application layer
- Evaluated models
- DeepSeek-V4-Pro, DeepSeek-V4-Flash
- DefensesMonitoring and auditing
- AttacksPoisoning and backdoors
- RiskDangerous agent actions
- ComponentMonitors
- DefensesarXiv 2607.25297
MTGuard: Hybrid Analysis for Secure MCP Tool Calls in LLM Agents
Proposes MTGuard, which blocks unsafe tool calls by comparing MCP call declarations with execution traces.
- arXiv
- 2607.25297
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- DeepSeek-V4-Flash, DeepSeek-V4-Pro, GPT-5.6 Luna
- DefensesMonitoring and auditing
- AttacksPoisoning and backdoors
- ComponentMCP and skills
- DefensesarXiv 2607.19913
Janus：面向长时程智能体安全的前瞻性潜在风险预测框架
提出 Janus 前瞻守卫，在长程智能体执行有害动作前拦截潜在风险
- arXiv
- 2607.19913
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- Vanguard
- DefensesMonitoring and auditing
- AttacksPrompt injection
- RiskDangerous agent actions
- ComponentMonitors
- DefensesarXiv 2606.22673
Agent MechSuits：面向多轮 CLI Agent 的机制子空间安全引导
提出 Agent MechSuits，对多轮 CLI Agent 做步级检测与子空间转向
- arXiv
- 2606.22673
- Published
- Layer
- Application layer
- Setting
- coding agent
- Evaluated models
- LLaMA-3.1-8B, Qwen-2.5-7B, Gemma-2-9B
- DefensesInference-time intervention
- AttacksMalicious use
- ComponentMonitors
Abstract白盒步级探针加稀疏转向可压低三模型恶意动作，同时提高生成崩溃率。
Agent MechSuits 是面向多轮 CLI 智能体的白盒、步级表示防御，并配有可执行轨迹基准 MAS。
- DefensesarXiv 2604.03870
Your Agent Is More Fragile Than You Think: Revealing Indirect Prompt Injection Vulnerabilities in Agentic LLMs
Using representation engineering to identify agent indirect injection hijacking before action submission
- arXiv
- 2604.03870
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- Qwen-2.5-14B, Qwen-2.5-32B, Qwen-3-4B and others9
- DefensesMonitoring and auditing
- AttacksPrompt injection
- ComponentMonitors
AbstractOn two Qwen models, all six defenses have Overall higher than no defense.
In the AgentDojo banking tool-calling environment, the authors examine whether indirect prompt injection causes open-source agents to execute attacker-specified actions, and the respective numbers for surface defenses and representation-based detection.
- DefensesarXiv 2602.20708
ICON：基于推理时纠正的 Agent 间接提示注入防御
Proposes inference-time defense ICON, which detects attention over-focusing and redirects to correct indirect prompt injection in agents.
- arXiv
- 2602.20708
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- Qwen-3-8B, Llama-3.1-8B, Mistral-8B and others6
- DefensesInference-time intervention
- AttacksPrompt injection
- ComponentMonitors
AbstractICON detects over-focusing and redirects at inference time to balance IPI safety and task utility.
ICON detects attention over-focusing from IPI at inference time and redirects, avoiding refusal that would terminate the agent trajectory.
- DefensesarXiv 2602.03085
Triggers in the Haystack: Extracting and Reconstructing LLM Backdoor Triggers
提出推理扫描器，在仅有模型文件时检测后门并重构触发词
- arXiv
- 2602.03085
- Published
- Layer
- Model layer
- Setting
- Models and APIs
- Evaluated models
- gemma-3-270m-it, DeepSeek-R1-Distill-Qwen-1.5B, Phi-4-mini-instruct and others8
- DefensesMonitoring and auditing
- AttacksPoisoning and backdoors
- ComponentFine-tuning and open weights
Abstract用记忆泄漏和内部信号重构触发词。
这是一个只做推理的 sleeper agent 扫描器：在不知触发词和目标行为时，检测因果语言模型里的条件后门并重构触发词。。问题在于，后门只在触发词出现时偏离基线，而开放仓库和网页投毒使检测方通常没有触发词、目标或激活语料。
- DefensesarXiv 2601.05755
VIGIL: defending LLM agents against tool-flow injection with "verify before submit"
提出 VIGIL 先验证后提交，防御 Agent 工具流注入
- arXiv
- 2601.05755
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- Qwen3-max, Gemini-2.5-pro
- DefensesInstruction–data separation
- AttacksPrompt injection
- ComponentMonitors
AbstractVIGIL 用意图约束下的提交前验证处理工具流注入。
VIGIL 用 verify-before-commit 处理智能体的工具流注入，并配有基准 SIREN。
You have reached the end