Flowise through 3.1.4 contains an insecure direct object reference vulnerability in the OpenAI...

CVE-2026-67622

Flowise through 3.1.4 contains an insecure direct object reference vulnerability in the OpenAI Assistants integration th

历史摘要 · 未重新核验 · GitHub Advisory

Ground Station prior to 0.6.0 contains an unauthenticated database-destruction and arbitrary-data...

CVE-2026-53984

Ground Station prior to 0.6.0 contains an unauthenticated database-destruction and arbitrary-data-injection vulnerabilit

历史摘要 · 未重新核验 · GitHub Advisory

Ground Station prior to 0.6.0 contains an unauthenticated blind server-side request forgery...

CVE-2026-53983

Ground Station prior to 0.6.0 contains an unauthenticated blind server-side request forgery vulnerability in the orbital

历史摘要 · 未重新核验 · GitHub Advisory

The System REST API accepts user-supplied file uploads without enforcing sufficient validation on...

CVE-2026-3418

The System REST API accepts user-supplied file uploads without enforcing sufficient validation on the file type or desti

历史摘要 · 未重新核验 · GitHub Advisory

Use after free in Skia in Google Chrome on Android prior to 151.0.7922.109 allowed a remote...

CVE-2026-19154

Use after free in Skia in Google Chrome on Android prior to 151.0.7922.109 allowed a remote attacker who had compromised

历史摘要 · 未重新核验 · GitHub Advisory

Out of bounds write in ANGLE in Google Chrome on Android prior to 151.0.7922.109 allowed a remote...

CVE-2026-19157

Out of bounds write in ANGLE in Google Chrome on Android prior to 151.0.7922.109 allowed a remote attacker to potentiall

历史摘要 · 未重新核验 · GitHub Advisory

Uninitialized Use in GPU in Google Chrome on Android prior to 151.0.7922.109 allowed a remote...

CVE-2026-19146

Uninitialized Use in GPU in Google Chrome on Android prior to 151.0.7922.109 allowed a remote attacker who had compromis

历史摘要 · 未重新核验 · GitHub Advisory

Insufficient validation of untrusted input in WebAPKs in Google Chrome on Android prior to 151.0...

CVE-2026-19143

Insufficient validation of untrusted input in WebAPKs in Google Chrome on Android prior to 151.0.7922.109 allowed a loca

历史摘要 · 未重新核验 · GitHub Advisory

league/commonmark: AttributesExtension href/src unsafe-link filter bypass via embedded control bytes

CVE-2026-71478

## Summary The `AttributesExtension`'s `href`/`src` unsafe-link filter (`AttributesHelper::filterAttributes()`) can be

历史摘要 · 未重新核验 · GitHub Advisory

Nx: Zip-Slip in the self-hosted remote cache

CVE-2026-71476

## Summary The Nx **self-hosted HTTP remote cache** extracts downloaded cache artifacts without constraining where file

历史摘要 · 未重新核验 · GitHub Advisory

Mermaid Architecture diagrams are vulnerable to prototype pollution

CVE-2026-71437

Rendering an untrusted `architecture-beta` diagram lets the diagram author write an arbitrary property with the value `h

历史摘要 · 未重新核验 · GitHub Advisory

Firewall for AI

Firewall for AI

历史摘要 · 未重新核验 · Akamai Security Research

PromptShield Home: Ambient Multimodal Prompt Injection Defense for Smart-Home Agents

随着多模态大语言模型（MLLM）被引入智能家居助手，智能体可直接感知视频和音频，这引入了新的安全问题：智能体能否区分真实用户指令与环境中看似指令的内容（如电视语音、屏幕文字、旁人的对话等）？本文提出 PromptShield-Home，一个

历史摘要 · 未重新核验 · arXiv cs.CR

When Does Latent Communication Pay? A Causal Audit of Relayed KV Caches in Multi-Agent LLMs

该论文针对多智能体大语言模型（LLM）系统中中继键值缓存（KV Cache）的因果效应进行了审计研究。在多智能体LLM系统中，智能体之间通常传递键值缓存而非显式文本，许多工作声称这种通信方式能够传递“潜在思维”（latent thought

历史摘要 · 未重新核验 · arXiv cs.CR

Combating Knowledge Corruption in Agent Systems: A Byzantine-Tolerant Secure Collaborative RAG Framework

检索增强生成(RAG)系统虽在一定程度上缓解了大型语言模型(LLM)的幻觉问题，但同时也引入了知识腐败攻击的新漏洞。攻击者可通过向RAG系统提供的文档中注入恶意内容(即文档投毒)来操纵LLM的输出，从而破坏系统的可靠性和安全性。针对这一威胁

历史摘要 · 未重新核验 · arXiv cs.CR

PriDyG: Privacy-preserving Dynamic Graph Inference with LLM-GNN Collaboration

本论文研究动态图上的边级差分隐私推理问题。在动态图中，关系数据不断更新，若反复用新数据训练模型，隐私损失会随更新次数累积，构成严重威胁。作者形式化定义了边级差分隐私动态图推理（EDG）问题，并提出 PriDyG 框架，将基于 GNN 的结构

历史摘要 · 未重新核验 · arXiv cs.CR

An Inline Control Architecture for Language Models in Intelligent Transportation Systems

本文面向智能交通系统（ITS）中车联网（V2X）场景下大语言模型（LLM）应用的语义安全防护问题。随着V2X系统在路侧单元和边缘节点中集成LLM用于消息摘要、操作员辅助和决策支持等非安全关键任务，传统的基于认证和消息完整性的安全机制无法覆盖

历史摘要 · 未重新核验 · arXiv cs.CR

ReputationChain: Robust Trust Updating for Blockchain-Enabled Supply Chains

本论文提出一个名为 ReputationChain 的参与者信任框架，旨在解决区块链供应链中仅依赖账本完整性无法判断参与者在未来风险敏感交易中是否值得信任的问题。现有声誉系统主要关注产品证据、全局反馈聚合或评论真实性，却较少关注重复双边通胀

历史摘要 · 未重新核验 · arXiv cs.CR

Blockchain Empowered Trustworthy Agent Networks: Foundations, Taxonomy, and Future Directions

本文是一篇关于区块链赋能可信智能体网络的综述与教程文章，由北京理工大学等机构的研究者撰写。文章首先回顾了从1980年代经典多智能体系统到2026年间智能体架构的演进，指出AI智能体正从孤立的任务执行器演变为能够通信、委派任务、调用工具、访问

历史摘要 · 未重新核验 · arXiv cs.CR

Server-side request forgery (ssrf) in Microsoft Office SharePoint allows an unauthorized attacker to perform spoofing over a network.

CVE-2026-70332

CVE-2026-70332 是 Microsoft Office SharePoint 中存在的一处服务端请求伪造（SSRF）漏洞。该漏洞允许未经授权的攻击者通过网络执行欺骗攻击（spoofing），可能进一步利用 SharePoint

历史摘要 · 未重新核验 · NVD