The silent Just-In-Time (JIT) provisioning feature in federated authentication implementations fails to properly segregate user roles during account creation when a federated user shares a username with a local user. This allows the provisioning process to overwrite existing roles of local users with roles assigned to the federated user. Exploitation requires a federated identity provider (IDP) w

Plack::Middleware::OAuth versions through 0.10 for Perl do not support the OAuth 2.0 state parameter. RequestTokenV2 builds the provider authorization redirect without issuing a state value, and AccessTokenV2 exchanges the callback code and registers the resulting token into the session (register_session) without verifying that the callback corresponds to an authorization request this session ini

Dancer2::Plugin::Auth::OAuth::Provider versions before 0.23 for Perl do not support the OAuth 2.0 state parameter. The authentication_url method builds the provider authorization redirect without issuing a state value, and the callback method exchanges the callback code and registers the resulting token into the session without verifying that the callback corresponds to an authorization request t

A security vulnerability has been detected in NousResearch hermes-agent up to 0.15.2. This affects the function DiscordAdapter._is_allowed_user of the file gateway/platforms/discord.py of the component Discord Platform Integration. Such manipulation leads to improper authentication. The attack can be launched remotely. This attack is characterized by high complexity. The exploitability is reported

A vulnerability was detected in NousResearch hermes-agent up to 2026.5.16. This impacts the function extract_media of the file gateway/platforms/base.py of the component Live Webhook Endpoint. Performing a manipulation results in path traversal. The attack may be initiated remotely. The exploit is now public and may be used. The vendor was contacted early about this disclosure but did not respond

A security flaw has been discovered in NousResearch hermes-agent up to 0.15.2. The affected element is the function shell.exec of the file tui_gateway/server.py. The manipulation results in protection mechanism failure. It is possible to launch the attack remotely. The exploit has been released to the public and may be used for attacks. The vendor was contacted early about this disclosure but did

A security vulnerability has been detected in NousResearch hermes-agent up to 2026.4.30. Affected is the function GatewayStreamConsumer._filter_and_accumulate of the file gateway/stream_consumer.py of the component Streaming Reasoning Tag Filter. The manipulation leads to improper handling of case sensitivity. The attack may be initiated remotely. The attack's complexity is rated as high. The expl

Improper access control in Microsoft Edge for Android allows an unauthorized attacker to bypass a security feature over a network.

Relative path traversal in Microsoft Edge for Android allows an unauthorized attacker to disclose information locally.

Absolute path traversal in Microsoft Edge for Android allows an unauthorized attacker to disclose information locally.

Time-of-check time-of-use (toctou) race condition in Microsoft Edge for Android allows an unauthorized attacker to execute code over a network.

Access of resource using incompatible type ('type confusion') in Microsoft Edge (Chromium-based) allows an unauthorized attacker to execute code over a network.

Ivanti Endpoint Manager Mobile代码注入漏洞

TOPSEC Firewalls - Remote Exploit (ELIGIBLEBACHELOR)

Fortigate Firewalls - Remote Code Execution (EGREGIOUSBLUNDER)

本论文回顾了广泛使用的网络入侵检测基准数据集CICIDS2017，对其数据收集流水线进行了深入分析，评估了数据集在流量生成、流构建、特征提取和标注方面的正确性、有效性和整体可用性。研究发现该数据集存在一系列严重问题：流量生成不符合实际网络环境、流构建存在错误、特征提取不一致、标签存在误标。这些问题严重影响了基于该数据集的机器学习模型的训练和评估。作者针对大部分缺陷提出了改进的数据处理方法，包括重新构建或重新标注了超过20%的原始流量记录。实验表明，在修正后的数据集上进行机器学习基准测试，模型性能得到显著提升，验证了数据质量问题对模型评估的巨大影响。本研究揭示了数据质量问题可能导致模型评估结果失真，并为类似数据集的构建和验证提供了预防建议。适合网络安全研究人员、数据科学家以及任何使用CICIDS2017进行入侵检测研究的人员阅读。

CrowdStrike与Zscaler宣布合作，将持续身份安全集成到零信任访问框架中。该合作旨在结合CrowdStrike Falcon平台的端点检测与响应（EDR）能力以及Zscaler的零信任网络访问（ZTNA）方案，实现基于用户身份、设备健康状态和实时威胁情报的动态访问控制。通过持续验证用户身份和端点安全状态，组织可以更有效地防止未经授权的访问，并降低凭据窃取和横向移动的风险。文章强调，传统的静态访问控制已不足以应对现代威胁，而持续身份安全能够根据风险变化实时调整权限，从而增强整体安全 posture。此合作适用于采用零信任架构的企业，特别是那些面临高级持续性威胁（APT）和勒索软件攻击的组织。

CrowdStrike 宣布推出面向 AI 代理的持续身份验证解决方案，旨在增强 AI 工作负载的安全性。该方案将身份安全与端点检测和响应（EDR）能力相结合，为 AI 代理提供实时身份验证和访问控制，防止未经授权的访问和恶意利用。公告未提及具体攻击事件或漏洞，而是聚焦于产品功能与安全架构。尽管标签包含 APT 和 Conti，但内容本身不涉及威胁活动或恶意软件。

Citrix NetScaler 内存泄漏(CVE-2025-5777)

Improper neutralization of input during web page generation ('cross-site scripting') in Microsoft Edge (Chromium-based) allows an unauthorized attacker to perform spoofing over a network.

Improper neutralization of input during web page generation ('cross-site scripting') in Microsoft Edge (Chromium-based) allows an unauthorized attacker to perform spoofing over a network.

Insufficient ui warning of dangerous operations in Microsoft Edge (Chromium-based) allows an unauthorized attacker to perform spoofing over a network.

Access of resource using incompatible type ('type confusion') in Microsoft Edge (Chromium-based) allows an unauthorized attacker to execute code over a network.

Improper input validation in Microsoft Edge (Chromium-based) allows an unauthorized attacker to execute code over a network.

Use after free in Microsoft Edge (Chromium-based) allows an unauthorized attacker to execute code over a network.

External control of file name or path in Microsoft Edge (Chromium-based) allows an unauthorized attacker to execute code over a network.

Improper access control in Microsoft Edge (Chromium-based) allows an unauthorized attacker to perform spoofing over a network.

Access of resource using incompatible type ('type confusion') in Microsoft Edge (Chromium-based) allows an unauthorized attacker to execute code over a network.

Operation on a resource after expiration or release in Microsoft Edge (Chromium-based) allows an unauthorized attacker to disclose information over a network.

Improper authorization in Microsoft Edge (Chromium-based) allows an unauthorized attacker to execute code over a network.

PLANET VDR-300NU ADSL Router - 未授权修改DNS

Acknowledgements

本文提出一种基于偏好优化的防御方法 SecAlign，用于缓解大型语言模型（LLM）面临的提示注入攻击。在 LLM 与外部数据源（如用户文档、网络检索结果、API 返回等）交互时，攻击者可将恶意提示注入这些外部内容，覆盖系统原有指令并执行恶意操作。SecAlign 的核心思路是构建一个偏好数据集，其中每个样本包含一个经过提示注入的输入、一个安全输出（响应合法指令）和一个不安全输出（响应注入指令）。然后利用偏好优化算法（如直接偏好优化 DPO）训练 LLM，使其更倾向于生成安全输出而非不安全输出。实验表明，SecAlign 能将多种提示注入攻击的成功率降至 10% 以下，即使面对训练中未见过的更复杂攻击也能保持有效，说明具有良好的泛化能力。同时，经过防御训练的模型在实用性上与原始模型相当，没有显著降低语言生成质量。该工作提供了首个已知的基于偏好优化的提示注入防御方法，代码已开源。

A vulnerability has been found in ForceInjection AI-fundermentals 2.0/3.0. Affected by this vulnerability is the function get_conversation_history of the file 08_agentic_system/memory/langchain/code/smart_customer_service.py of the component Memory Recall Handler. The manipulation leads to use of weak hash. Remote exploitation of the attack is possible. A high degree of complexity is needed for th

A vulnerability has been found in FederatedAI FATE up to 2.2.0. This affects the function QueuePushReqStreamObserver.initEggroll of the file java/osx/osx-broker/src/main/java/org/fedai/osx/broker/grpc/QueuePushReqStreamObserver.java of the component OSX Broker. Such manipulation of the argument rollSiteSessionId/dstRole/dstPartyId leads to exposure of data element to wrong session. The attack can

Akamai 发布了面向 AI 工作负载的专用防火墙产品“Firewall for AI”，旨在保护 AI 推理端点免受 DDoS 攻击、botnet 流量和其他恶意请求的侵害。该产品基于 Akamai 的全球分布式边缘网络，提供实时的流量监控和威胁缓解能力。文章指出，随着 AI 模型的广泛部署，针对 AI 基础设施的攻击面不断扩大，特别是大规模的 HTTP DDoS 攻击和针对模型端点的滥用行为。Firewall for AI 能够识别并过滤恶意流量，同时不影响正常用户的请求。该产品还集成了威胁情报，可以自动更新规则以应对新兴的 botnet 和 DDoS 攻击模式。尽管文章中未提及具体的攻击事件或漏洞，但这一产品反映了 AI 安全防御领域的发展趋势。

Firefox browsers

Sign in to Cloud

Sign Up for Free Cloud Tier

iOS 26.4.2 and iPadOS 26.4.2

iOS 18.7.8 and iPadOS 18.7.8

iOS 18.7.7 and iPadOS 18.7.7

iOS 26.4 and iPadOS 26.4

Background Security Improvements for iOS, iPadOS, and macOS

iOS 16.7.15 and iPadOS 16.7.15

iOS 15.8.7 and iPadOS 15.8.7

iOS 26.3 and iPadOS 26.3

iOS 18.7.5 and iPadOS 18.7.5

iOS 26.2 and iPadOS 26.2

iOS 18.7.3 and iPadOS 18.7.3

Compressor 4.11.1

iOS 18.7.2 and iPadOS 18.7.2

iOS 26.1 and iPadOS 26.1

iOS 26.0.1 and iPadOS 26.0.1

iOS 18.7.1 and iPadOS 18.7.1

Android Code Search

Android Devices

Secure an Android device

Android 16 QPR2

Mobile network security

iOS 26.5 and iPadOS 26.5

iOS 18.7.9 and iPadOS 18.7.9

iOS 16.7.16 and iPadOS 16.7.16

iOS 15.8.8 and iPadOS 15.8.8

MFSA 2026-49Security Vulnerabilities fixed in Firefox for iOS 151.0

MFSA 2026-52Security Vulnerabilities fixed in Firefox for iOS 151.1

MFSA 2026-53Security Vulnerabilities fixed in Firefox for iOS 151.2

MFSA 2026-55Security Vulnerabilities fixed in Focus for iOS / Klar 151.3.1

MFSA 2026-56Security Vulnerabilities fixed in Firefox for iOS 152.0

【Cloud Mate】产品退市及停止服务公告

iOS 26.5.2 and iPadOS 26.5.2

随着大语言模型（LLM）和智能体系统被集成到实际应用中，确保其安全性至关重要。护栏系统作为AI安全的核心组件，能够检测并阻止发往LLM的恶意指令。然而，对生产AI系统进行黑盒对抗模拟的研究人员往往难以区分是护栏拦截还是LLM自身拒绝。这一区分至关重要，因为绕过护栏和绕过LLM安全对齐的技术差异显著，直接影响攻击技术的选择与优化。本文提出了首个黑盒护栏侦察方法，仅通过HTTP、词法和时序信号的行为监控来检测目标AI系统中护栏的存在，假设仅具备黑盒访问权限且对护栏或AI系统零先验知识。实验表明，该方法能以100%的准确率检测护栏存在，且良性交互与恶意交互在行为统计上具有显著差异（q<0.001）。进一步地，该方法能识别护栏设计拦截的内容类别，并在未见提示上以平均F1分数98%区分护栏拦截与LLM拒绝。该方法对攻击者评估AI系统防御、红队测试及防御方评估自身护栏有效性均具有重要意义。

智能合约漏洞主要是逻辑漏洞，其检测需要结构化的、逐步的攻击模式与合约语义的程序性知识。现有基于LLM的方法难以自动生成此类知识：基于提示的方法依赖人工设计的检测规则，而微调则需要大量标注样本，这在智能合约领域极为稀缺。本文提出EvoVuln，一个自动化框架，将漏洞检测重新定义为程序性知识演化问题，仅需最少量的标注样本即可合成并优化检测逻辑。EvoVuln引入两个关键机制：一是基于控制反转（IoC）架构的运行时，将检测规则编译为可执行策略，严格解耦确定性控制流与LLM语义推理，确保逻辑忠实执行并生成密集的诊断遥测用于精确定位错误；二是两阶段演化流水线，通过溯因语义调试打磨规则而无需任何参数更新：冷启动阶段利用自动合成的边界案例对初始规则进行引导和压力测试，少样本演化阶段则利用每个漏洞类型仅5个漏洞和5个安全样本将策略锚定到真实世界语义。在五种真实漏洞类型上的评估表明，EvoVuln达到71%的宏观平均F1分数，优于所有基线。演化的程序性知识可跨模型迁移：它使轻量低成本的模型超越更大的零样本模型19个百分点，且可无再训练地迁移到其他LLM，单次演化成本低于50美元。

本文提出 AgentFlow，首个针对 LLM 代理程序的静态分析框架。LLM 代理程序通常基于代理框架开发，其行为不仅依赖传统控制流和数据流，还受代理依赖关系影响，如模型、提示、工具、记忆及多代理编排逻辑等。这些依赖关系往往通过框架语义（如构造函数、工具装饰器、代理交接声明）表达，现有静态分析工具难以恢复。AgentFlow 构建了代理依赖图（ADG），一种框架无关的图表示，将代理、提示、模型、能力、记忆状态和控制策略作为类型节点，并将组件依赖、控制流和数据流依赖作为类型边。基于 ADG，AgentFlow 支持代理治理和安全分析，包括代理物料清单（BOM）生成和提示到工具风险检测。作者针对五个代表性代理框架实现了 AgentFlow，并在包含 5,399 个真实代理程序的 AgentZoo 语料库上评估。结果表明，AgentFlow 比现有基于 AST 的静态分析工具恢复更丰富的代理实体和依赖关系，生成更多依赖感知的代理 BOM，并在真实代理程序中发现 238 个污点类型的提示到工具风险。这些结果证明 ADG 为理解、治理和保护新兴代理软件提供了实用基础。

本论文提出了一种名为“认知防火墙”（Cognitive Firewall）的主动式运行时监督框架，旨在增强大型语言模型（LLM）的安全性。当前主流的运行时安全措施通常将用户提示或模型响应作为孤立消息进行评估，这导致它们难以检测跨多轮对话中积累的恶意意图、验证声称的权限，或识别被分解到多次交互中的有害目标。认知防火墙通过在用户与受保护的目标模型之间插入一个独立的监督模型，将安全评估分解为四个分类门控：（1）意图门（Intent Gate）——识别请求的操作目标；（2）零信任上下文门（Zero Trust Context Gate）——将声称的角色和权限视为未经验证的证据；（3）一致性门（Consistency Gate）——检测跨对话轮次的升级和分解行为；（4）输出风险门（Output Risk Gate）——在发布前检查候选响应。这些门控的决策通过升级机制而非分数平均进行组合，使得任何可信的危险信号都能阻止交互，同时保留可审计的推理过程。实验在四个越狱基准测试和一个良性安全测试集上进行，结果表明认知防火墙显著降低了单轮、多轮、基于权限和人工制作的攻击的成功率。在三个攻击集上，攻击成功率降至2%或以下；在最困难的人工制作攻击集上降至14%，同时保持了8%的过度拒绝率。这些结果说明，分解的、对话级别的监督可以提高LLM安全的主动防御和可审计性。

Cloud & Application Security

From Scanner to Stealer: Inside the trivy-action Supply Chain Compromise

Next-Gen Identity Security

CrowdStrike Named an Innovation and Growth Leader in the 2026 Frost Radar™: Cloud and Application Runtime Security

CrowdStrike Expands Identity Leadership with OpenID and IDPro

cloud architecture assessment

ransomware and multifaceted extortion defense

Google Cloud Next 2026Catch-up sessions on the keynotes and select sessions are now available on demand.Explore content on-demand

The Identity Problem Hiding in AI Agent Deployments

94% of Organizations Report Cloud Breaches: CrowdStrike State of CDR Survey

Falcon Cloud Security June 2026 Release: Updates for Azure and Google Cloud

零配置协议自1999年提出以来，为网络设备提供了即插即用的连接机制，无需手动配置DNS或DHCP服务器。如今，几乎所有个人电脑、打印机、扫描仪等设备都默认启用了这类协议，用户往往毫不知情。物联网生态的蓬勃发展更是高度依赖这些协议（如mDNS、LLMNR、DHCPv6等）。然而，这些协议在设计上存在多个安全缺陷，使得攻击者可以实施多种中间人攻击（MITM）。尽管此前已有少量研究提及个别漏洞，但整体威胁被严重低估。本文对这些设计缺陷进行了系统性的深度剖析，覆盖了多类零配置协议的攻击面。作者开展了300次实验，测试了不同厂商实现的安全性，并基于实验结果揭示了问题的广泛性。作为防御手段，作者提出了利用Zeek（原名Bro）入侵检测框架来检测此类攻击的方法，并公开了攻击代码和Zeek检测脚本。这些代码以可复现研究结果的方式提供，但设计上避免被脚本小子直接利用。本文旨在提升安全社区对零配置协议威胁的认知，并给出实用的检测方案。适合网络管理员、安全运维人员以及物联网安全研究者阅读。

该论文研究了Android设备中基于功率的侧信道攻击。现代计算机系统面临软件功率侧信道攻击的威胁，攻击者可以通过分析功耗变化窃取敏感信息。以往攻击如PLATYPUS依赖专用接口的直接功率信号，Hertzbleed则利用功率相关的时序变化。但随着平台对直接功率信号的访问限制，探索其他功率相关信号变得重要。本文发现Android移动设备通过传感器框架暴露了多种功率相关信号，这些信号反映了功耗的寄生影响。作者系统分析了多个设备上Android传感器框架中非特权传感器，识别出新的泄露原语：地磁旋转矢量传感器的旋转依赖功率泄露。通过对比地面真值，部分传感器的相关性超过0.9。极端情况下，CPU压力可使指南针指针偏转约30度。论文通过两个案例研究评估了泄露原语：远程攻击者通过Google Chrome浏览器、本地攻击者通过安装的App。特别地，作者实现了端到端的像素窃取攻击，以每像素5-10秒的速率绕过浏览器的跨源隔离。此外，利用新泄露原语展示概念验证的AES攻击，泄露单个密钥字节。论文揭示了Android传感器框架的安全风险，为防御者提供了新的关注点。

该研究针对人工耳蜗（CI）用户这一特殊人群，系统评估了音频深度伪造（audio deepfakes）对其语音识别与真实性判断的影响。研究设计了包含真实语音与深度伪造语音的听感实验，招募了正常听力与CI用户两组受试者，测量他们在不同噪声环境下的辨别准确率、反应时间和主观信心。实验发现，与正常听力者相比，CI用户在区分真实与伪造语音方面的整体准确率显著更低，且个体间差异极大；深度伪造音频中细微的频谱畸变在CI用户中往往被忽略或被错误解释为环境噪声。研究还通过声学分析揭示了CI信号处理链（如滤除高频、压缩动态范围）对深度伪造痕迹的掩蔽效应，进而导致检测能力下降。该工作首次量化了听觉辅助设备对深度伪造检测的负面影响，指出当前深度伪造检测模型未考虑听障用户的感知局限性，并建议在AI合成语音安全评估中纳入辅助设备用户的视角。研究最后提出了面向CI用户的个性化语音真实性提示、可调节的音频保真度指标等缓解方向。

该论文系统性地研究了视频生成模型（VGM）产生不安全内容（如暴力、恐怖、色情、政治敏感等）的风险。作者首先从4chan和Lexica等来源收集可能诱导不安全内容生成的提示词，并使用三种开源最先进视频扩散模型（如VideoFusion、ModelScope等）生成视频，经过滤后从5607个原始视频中筛选出2112个候选不安全视频。通过聚类和主题编码，归纳出五类不安全视频：扭曲/怪异、恐怖、色情、暴力/血腥和政治。经IRB批准，招募403名在线参与者对视频进行标注，最终确认937个不安全视频，并构建了首个VGM生成不安全视频数据集。针对现有防御方法（输入过滤或输出过滤）的不足，论文提出一种名为潜在变量防御（LVD）的新方法，该方法在模型内部采样过程的初始阶段检测不安全样本，通过分析潜在空间中的中间表示来判断是否生成不安全内容。实验表明，LVD在三个开源模型上分别达到0.99、0.92和0.91的防御准确率，且在对抗性提示和图像到视频扩散模型上准确率均超过0.90。与基线方法相比，LVD在大规模采样时可将时间和计算资源降低10倍。此外，LVD还能与其他防御方法组合使用以提升整体性能。论文将公开数据集和代码。

本文针对近年流行的扩散模型（diffusion models）提出了一种新的黑盒成员推理攻击（Membership Inference Attack, MIA）框架。随着预训练扩散模型的高质量发布，越来越多的用户下载这些模型并使用下游数据集进行微调，这带来了显著的隐私泄露风险。作者首次基于分数（scores-based）设计了适用于扩散模型的成员推理攻击，并在更严格的黑盒访问设置下进行。该框架考虑了四种不同的攻击场景和三种攻击类型，能够针对任何流行的条件生成器模型，实现了高达0.95的AUC（Area Under Curve）精度。代码已开源。研究揭示了扩散模型在下游任务微调中的隐私脆弱性，为后续防御研究提供了基准。

该论文聚焦于联邦学习（FL）中的隐私保护问题。联邦学习作为一种分布式机器学习范式，允许多个客户端在不共享原始数据的前提下，通过聚合服务器联合训练全局模型。然而，现有研究表明，恶意聚合服务器可能通过客户端上传的模型更新发起推理攻击，窃取客户端的隐私信息。安全聚合（Secure Aggregation）技术旨在让聚合服务器仅能获取更新之和或平均值，从而保护个体更新。但现有的安全聚合协议存在通信开销大、计算复杂、轮次过多等缺陷。本文提出SAFELearn，一个通用的高效隐私联邦学习系统设计，专门抵御针对个体模型更新的推理攻击。其核心思想是结合安全多方计算（MPC）或全同态加密（FHE）技术，在不依赖可信第三方的前提下，实现仅两轮通信的聚合，且客户端无需执行昂贵的密码操作，同时支持客户端掉线处理。作者基于两方安全计算实例化该系统，并在实验中证明：聚合500个参数超过30万的模型，耗时不到0.5秒。总体来说，SAFELearn在保证隐私性的同时，大幅提升了联邦学习中安全聚合的效率与实用性。

本文针对现代Web应用的黑盒漏洞扫描面临的挑战展开研究。现代Web应用具有复杂性、动态性及对第三方组件的依赖，使得漏洞检测愈发困难。作者指出黑盒扫描的两个关键难点：一是实现深度渗透需要模拟用户交互和异步行为（如点击、表单提交、AJAX请求）；二是检测存储型XSS等复杂注入攻击需要发现页面间的数据依赖关系。基于这些挑战，论文提出了三个核心支柱：导航建模（正确描述应用状态和转换）、遍历（高效探索状态空间）以及跟踪跨状态依赖（捕获数据在页面间的流动）。以往工作通常只关注其中一个方面，而本文提出的Black Widow方法则综合了这三者，实现了一种数据驱动的黑盒爬虫与扫描框架。Black Widow通过动态分析页面中的表格、链接、表单等元素，自动构建导航模型，并利用依赖跟踪机制识别从用户输入到输出之间的数据流，从而有效发现存储型XSS漏洞。实验表明，在多个真实Web应用上，Black Widow的代码覆盖率相比其他爬虫提升了63%至280%；在漏洞发现方面，零误报且比现有方法发现了更多的XSS漏洞。此外，在旧版应用上，Black Widow也发现了被其他方法遗漏的漏洞，并在生产环境的软件（如HotCRP、osCommerce、PrestaShop和WordPress）中发现了新漏洞。本文适合Web安全研究人员、渗透测试工具开发者以及关注自动化漏洞检测的读者阅读。

本文重新审视了网络设备中RSA弱密钥的普遍性问题。自2012年起，多个学术研究揭示了互联网HTTPS主机中RSA密钥的漏洞，指出弱密钥（例如共享素数、因子可分解等）广泛存在，并归因于低熵随机数生成器、厂商实现缺陷等因素。2016年的一项后续研究对大量RSA模数进行长期分析，评估了厂商和终端用户针对原始披露所采取的行动，结果发现弱密钥仍十分普遍，厂商几乎未采取任何补救措施。本文在前述研究基础上进行了再评估，重点聚焦于更大尺寸的RSA模数（2048位），因为随着计算能力提升，512位和1024位密钥逐渐被淘汰，而2048位已成为当前主流。作者收集并分析了大规模2048位RSA模数数据集，重现了已知的弱密钥类型（如共享因子、低熵生成等），并进一步发现了一些之前未被详细记录的新异常模式。研究结果表明，尽管行业已提高密钥长度标准，但弱密钥问题仍未根除，在2048位密钥中仍然存在一定比例的可攻击样本。本文的核心贡献在于：1）对2048位RSA弱密钥进行了系统性的重新评估，更新了2016年以来的数据；2）揭示了新的异常类型，为后续漏洞检测和厂商修复提供了依据；3）强调了网络设备制造商在随机数生成和密钥生命周期管理方面仍需改进。适合安全研究人员、网络设备厂商及负责HTTPS基础设施运营的工程师阅读。

本文针对k匿名化在数据脱敏中的隐私风险进行了深入研究。尽管k匿名化因简单、符合法规且保持数据可用性而被广泛采用，但隐私界长期批评其保护不足。非专业人员常认为在没有辅助信息的情况下，k匿名化无法被攻破。本文首次提出组合精炼攻击（Combinatorial Refinement Attacks, CRA），这是一种新型隐私攻击方法，专门针对使用局部编码（local recoding）生成的k匿名化数据集。CRA不依赖任何外部辅助信息或对数据分布的假设。它利用ARX（临床数据匿名化广泛使用的开源软件）中局部编码的效用优化行为，构建线性规划模型，显著减少可能的敏感值空间。为验证该攻击的有效性，作者与一个免费社区健康诊所网络合作，该环境既缺乏辅助信息又因资源和法规限制而采用开源k匿名化方案。基于真实临床微观数据的实验表明，即使没有外部信息，现有的匿名化框架也无法提供所承诺的隐私保护水平，引发了严重的隐私担忧。本文的核心贡献是揭示k匿名化在无辅助信息场景下的脆弱性，并提供了可复现的攻击方法。适合数据隐私研究人员、医疗数据管理者及匿名化工具开发者阅读。

本文针对RNS-CKKS全同态加密（FHE）方案中缩放因子与模数刚性耦合的问题，提出了一种名为“嫁接”（Grafting）的解耦技术。在标准RNS-CKKS中，缩放因子决定数值精度，模数保证安全性，但两者必须紧密关联，导致设计受限，例如难以找到合适的NTT素数，且精度灵活性不足。本文的核心方法是通过重新设计密钥生成、加密和计算流程，允许在加密过程中独立设置缩放因子，从而打破耦合。具体地，Grafting技术引入额外的转换步骤，使得用户可以在保持相同模数的情况下调整缩放因子，或者在不同缩放因子之间进行切换。实验表明，该方法在不降低安全性的前提下，显著提升了参数选择的灵活性，并减少了因参数不匹配导致的效率损失。主要贡献包括：形式化地定义了缩放因子与模数解耦的必要条件；提出了Grafting算法的具体实现；通过理论分析和实验验证了其正确性与性能优势。本文适合FHE研究人员、密码学工程师以及需要优化同态加密性能的开发人员阅读。

本研究针对企业董事会在网络安全风险决策中的实践进行了定性分析。通过深度访谈和案例研究，作者揭示了董事会成员在评估网络威胁、分配资源以及制定安全策略时的真实行为与认知偏差。研究发现，多数董事会依赖有限的信息来源（如管理层简报或外部审计报告），缺乏对技术细节的理解，导致决策往往基于直觉而非数据。此外，组织文化、激励机制和沟通障碍显著影响决策质量。研究提出了一个决策框架，帮助董事会结构化地考虑风险偏好、合规要求和业务连续性。该工作为安全领导者提供了与高层沟通的实证基础，并为改进公司治理中的安全监督指明了方向。

该论文研究了资源公钥基础设施（RPKI）及其路由来源验证（ROV）在部署中面临的关键障碍。尽管RPKI旨在通过绑定IP前缀与授权自治系统（AS）来防止前缀劫持，全球仍有超过40%的路由未受RPKI保护。许多网络运营商因担心ROV判定为无效的路由可能是合法的，而拒绝部署RPKI，因为过滤这些路由可能损害网络服务质量，尤其是影响网络连通性。论文的核心贡献在于提出了一种评估无效路由合法性的方法，旨在区分真正的劫持与因配置错误等原因导致的合法路由，从而降低运营商的部署顾虑。实验可能基于真实路由数据验证方法的有效性。该研究适合路由安全、网络运营和研究领域的读者。

本文提出了一种相关感知的安全排序与排列方法，旨在提升迭代式两方图分析中安全多方计算的效率。在联合图分析场景中，双方需要在不泄露原始数据的情况下处理包含丰富拓扑和属性信息的图数据。现有方法存在冗余的安全排序或高开销的安全洗牌操作，导致效率低下。部分工作通过放松安全假设（如引入差分隐私或辅助方）来改进，但可能牺牲隐私保护强度。本文提出的方法利用图分析中数据相关性特征，优化了安全排序和排列协议，减少了不必要的计算和通信开销。实验表明，该方法在保持同等安全级别的前提下，显著提升了迭代图分析任务（如PageRank、连通分量等）的执行效率。该研究适合安全计算和图分析领域的研究人员阅读。

无人机开源飞控 PX4-Autopilot堆栈溢出漏洞 CVE-2025-15150

MongoDB内存泄漏 CVE-2025-14847(MongoBleed)

1Panel 代理证书验证绕过导致任意命令执行漏洞(CVE-2025-54424)

ModelContext Inspector 未授权访问漏洞(CVE-2025-49596)

llamaindex SQL 注入漏洞(CVE-2025-1750)

Screen 本地权限提升漏洞(CVE-2025-23395)

Zabbix认证后SQL注入漏洞(CVE-2024-42327)

ProjectSend认证绕过漏洞(CVE-2024-11680)

PyTorch库RPC框架反序列化RCE漏洞(CVE-2024-48063)

Grafana认证后DuckDB-SQL注入漏洞(CVE-2024-9264)

SPIP BigUp Unauthenticated RCE(CVE-2024-8517)

Rejetto HFS 远程命令执行漏洞(CVE-2024-39943)

Apache HugeGraph-Server Command Execution In Gremlin（CVE-2024-27348）

Zabbix 后台延时注入(CVE-2024-22120)

CrushFTP 认证绕过漏洞（CVE-2024-4040）

Alert for CVE-2026-21992

Alert for CVE-2025-61884

Alert for CVE-2025-61882

Alert for CVE-2024-21287

Alert for CVE-2022-21500

Alert for CVE-2021-44228

Linux 内核提权 CVE-2026-31431(copy-fail)

Alert for CVE-2026-35273

【安全通告】Gitea act_runner 容器逃逸漏洞风险通告（CVE-2026-58053）

本文是一篇系统化知识（SoK）论文，旨在为网络安全事件响应领域的影响因素提供一个统一的分类框架。尽管事件响应已成为研究者和从业者关注的关键领域，但现有文献缺乏一个系统化的组织框架来整合跨技术、人机交互、组织理论和人为因素等多领域的知识。作者通过系统性文献综述，分析了417篇学术文献和40篇非科学出版物，从中提取了影响组织网络安全事件准备和响应的因素，并构建了“网络安全事件响应影响因素分类法”（CIR-IF Taxonomy）。该分类法将现有的实证发现进行分类，提供了从1999年至2024年中期的全面知识概览。此外，作者将CIR-IF Taxonomy与七个已有的科学框架以及NIST网络安全框架（NIST SP 800-61r3事件响应概要）中的元素进行了系统比较。比较结果表明，CIR-IF Taxonomy在驱动和塑造事件响应的因素方面提供了更丰富、更严谨且更系统化的视角。该研究适合事件响应研究人员、安全运营团队以及希望系统化理解事件响应影响因素的组织阅读。

本文提出了一种名为 ElephantAgent 的协议，旨在解决智能体系统（Agentic Systems）中因依赖外部工具和持久化记忆而引入的上下文状态中毒攻击问题。当前研究表明，恶意构造的工具描述和中毒记忆可以隐蔽地影响智能体行为。作者指出，这反映了智能体系统缺乏可验证的上下文状态连续性。ElephantAgent 借鉴了先前状态连续性机制（如 Nimble）的思想，将其扩展到智能体系统的演化上下文状态。文中定义了上下文状态作为智能体整体上下文中边界安全关键的子集（例如工具状态和记忆）。在处理每个查询之前，ElephantAgent 重新计算本地上下文状态的摘要，并与最新的授权摘要进行比对。通过使用复制的可信硬件，ElephantAgent 维护了一个可线性化的授权上下文状态转换账本，并检测带外状态篡改。为了处理带内语义滥用，ElephantAgent 还提供了历史可追溯性，支持有条件的后验审计和恢复到已知良好的先前状态。实验部分（摘要未详细说明）应验证了该协议的有效性。本文适合对 LLM 安全、智能体系统安全、可审计性感兴趣的研究人员阅读。

硬件安全验证是一个多阶段流程，工程师需要处理复杂的设计分析、威胁考量与验证策略，但现有验证环境缺乏结构化的安全指导支持。虽然对话式AI可以提供按需帮助，但直接使用通用型聊天机器人（如ChatGPT、Gemini）存在风险，因为它们容易产生幻觉且依赖静态过时知识。为此，本文提出VeriChat——一个面向硬件安全验证的领域专用对话助手，旨在增强而非取代现有验证工作流。VeriChat采用检索增强的多智能体架构，包含三个专门智能体，它们协作降低幻觉，提升响应的透明性与可靠性。除问答功能外，VeriChat还集成了开源EDA工具（Icarus Verilog、Yosys、SymbiYosys），能直接对用户提供的RTL设计执行语法检查、综合分析、仿真与形式化验证。综合评估显示，VeriChat的忠实度得分为87.73%，显著优于主流商用模型。通过一个AES S-Box IP上的硬件木马检测案例，VeriChat在多次对话中自主完成木马识别、仿真与形式化证明，成功发现隐蔽的密钥泄露漏洞。

该论文研究了轻量级机器学习模型在工业物联网（IIoT）入侵检测中的跨域泛化失败问题。虽然轻量级模型因其适合资源受限的边缘部署而被广泛提出，但大多数报告的结果仅在训练网络内评估，未验证其在未见网络上的表现。本研究在一种IIoT数据集上训练了四种轻量级架构，并使用跨三个数据源共有的特征表示，在另外两个结构不同的IIoT数据集上进行零样本评估。对两个最优模型的解释性分析显示，它们都过度依赖粗粒度的端口类别特征：最影响决策的端口类别在源域攻击流量中的出现频率是目标域的96至435倍，这表明粗化端口分辨率只是将已知的捷径（shortcut）从一个位置移动到另一个位置。此外，在自然不平衡的类别分布下，评估协议的不同选择会逆转两个目标域中哪个域泛化挑战更大的结论。研究还评估了对抗鲁棒性和通过有限目标域暴露的恢复能力：对抗扰动的鲁棒性与跨网络泛化无关，而通过自适应恢复的效果因架构而异。这些发现表明，部署准备度应基于真实类别分布下的跨网络评估，而非仅依赖域内准确率。

这篇综述系统地梳理了身份证件伪造攻击与检测方法的最新进展。随着生成式AI工具的普及，伪造能力发生了根本性转变：非专业人员也能借助AI生成高逼真度的证件或对现有证件进行字段级篡改，而现有的检测基准尚未跟上这一威胁。论文首次将物理呈现攻击（如印刷品翻拍）、数字注入攻击（如向数字验证流程插入伪造图像）和完全由生成式AI驱动的合成攻击统一在一个身份证件验证威胁模型下，揭示了不同攻击类型特有的取证失效模式。检测方法方面，论文从规则启发式方法、取证定位、注入感知流水线，一直追踪到大模型和少样本框架。通过对2019-2025年间公开数据集的系统审计，作者发现基准测试与真实部署之间存在持续的“现实差距”。此外，论文分析了大型多模态模型在证件篡改中的表现，识别出“脚本相关生成不稳定性”（SDGI）——一种非拉丁字母文本修补中反复出现的排版失败模式。最后，在未见过的合成身份证上的零样本基准测试显示，即使最强的公开模型在安全导向的操作条件下APCER值仍超过25%，暴露出跨域泛化的重大局限。论文最后指出了未来方向：构建基于取证、保护隐私并具有法律问责性的身份验证系统。

The silent Just-In-Time (JIT) provisioning feature in federated authentication implementations fails to properly segregate user roles during account creation when a federated user shares a username with a local user. This allows the provisioning process to overwrite existing roles of local users with roles assigned to the federated user. Exploitation requires a federated identity provider (IDP) w

Dancer2::Plugin::Auth::OAuth::Provider versions before 0.23 for Perl do not support the OAuth 2.0 state parameter. The authentication_url method builds the provider authorization redirect without issuing a state value, and the callback method exchanges the callback code and registers the resulting token into the session without verifying that the callback corresponds to an authorization request t

Plack::Middleware::OAuth versions through 0.10 for Perl do not support the OAuth 2.0 state parameter. RequestTokenV2 builds the provider authorization redirect without issuing a state value, and AccessTokenV2 exchanges the callback code and registers the resulting token into the session (register_session) without verifying that the callback corresponds to an authorization request this session ini

A vulnerability was detected in NousResearch hermes-agent up to 2026.5.16. This impacts the function extract_media of the file gateway/platforms/base.py of the component Live Webhook Endpoint. Performing a manipulation results in path traversal. The attack may be initiated remotely. The exploit is now public and may be used. The vendor was contacted early about this disclosure but did not respond

A security vulnerability has been detected in NousResearch hermes-agent up to 0.15.2. This affects the function DiscordAdapter._is_allowed_user of the file gateway/platforms/discord.py of the component Discord Platform Integration. Such manipulation leads to improper authentication. The attack can be launched remotely. This attack is characterized by high complexity. The exploitability is reported

A security flaw has been discovered in NousResearch hermes-agent up to 0.15.2. The affected element is the function shell.exec of the file tui_gateway/server.py. The manipulation results in protection mechanism failure. It is possible to launch the attack remotely. The exploit has been released to the public and may be used for attacks. The vendor was contacted early about this disclosure but did

Android 兼容性测试套件（CTS）是用于验证 Android 设备是否符合 Android 兼容性定义文档（CDD）要求的工具集。CTS 包含一系列测试用例，覆盖了 Android 框架的各个方面，如 API 兼容性、权限模型、安全性、性能等。设备制造商需通过 CTS 测试，确保其设备能够正常运行 Android 应用并保障用户体验的一致性。CTS 并非安全漏洞修复公告，而是 Android 生态系统质量保证的重要组成部分。定期运行 CTS 有助于发现设备实现与 Android 标准之间的差异，从而提前规避潜在兼容性和安全问题。

该公告来源于 Android 安全公告，但内容实为 Android 兼容性定义文档（CDD），而非安全漏洞公告。CDD 是 Android 平台用于定义设备兼容性要求的官方文档，规定了硬件、软件、API 等方面的标准，以确保应用在不同 Android 设备上的一致性。本次发布可能涉及对兼容性要求的更新或澄清，但未提及任何安全漏洞或 CVE。因此，本报告不涉及具体的漏洞细节或攻击风险。

该输入是Android安全公告网站上关于工具、构建和相关参考的页面。此页面主要提供Android平台开发中使用的构建工具链、编译系统及相关参考文档的说明，不包含任何安全漏洞或修复信息。作为参考资源，它旨在帮助开发者理解Android构建流程和工具使用，并非安全公告。因此，没有具体的漏洞需要评估。

此公告是 Android 安全公告发布的最新版《兼容性定义文档》（CDD）。CDD 是 Android 生态系统的核心规范，它定义了设备必须满足的硬件和软件要求，以确保能够运行 Android 应用程序并兼容 Google 服务。文档涵盖安全模型、权限模型、加密要求、更新机制等关键安全相关条款。虽然 CDD 本身不直接修复安全漏洞，但遵守 CDD 是设备获得 Google 认证的前提，有助于确保设备具备一致的安全基线。对于防御者和设备制造商而言，定期审查并遵循最新 CDD 版本是维护 Android 设备安全性的重要环节。本公告不包含特定 CVE 或漏洞详情，严重性未明确评分。

Android 安全公告（Android Security Bulletin）发布了最新一期安全更新。公告涵盖了 Android 框架、系统组件、内核等多个层面的安全修复。由于公告未提供具体的 CVE 编号或漏洞详情，当前无法确认涉及的具体漏洞类型及影响范围。通常，Android 安全公告会修复多个高危漏洞，包括远程代码执行、权限提升等。建议用户关注官方渠道获取详细补丁信息，并及时更新设备至最新的安全补丁级别。

Apple 安全发布页面发布了一则公告，标题为“update the software on your Mac”，呼吁用户更新 Mac 上的软件。公告可能涉及 macOS 或其他 Apple 软件的安全修复，但未提供具体的漏洞详情、CVE 编号或受影响版本。作为防御方，应关注官方公告以获取更多细节。

Apple 于 2026 年 7 月 5 日发布安全公告，建议用户更新 iPhone 或 iPad 上的软件。此次更新旨在修复可能存在的安全漏洞，以保护设备免受潜在威胁。虽然公告未提供具体漏洞详情或 CVE 编号，但 Apple 定期发布此类更新以增强系统安全性。建议用户尽快安装最新版本。

Apple Security Releases 发布了标题为“submit your research”的安全公告。该公告可能涉及安全研究提交相关的内容，但公告正文摘要未提供，因此无法获取具体的漏洞细节、受影响的组件、攻击向量或影响范围。建议安全团队关注 Apple 官方更新的发布，以便及时获取详细信息并采取相应措施。

该页面是苹果官方安全公告的帮助页面，标题为“Get help with security issues”。页面未提供具体的漏洞描述、受影响的版本或CVE编号。它可能包含指向如何报告安全问题、获取安全建议或查看历史公告的通用指导。没有实质性的漏洞技术细节。

Apple 发布了安全改进公告（Background Security Improvements），该公告于 2026 年 7 月 5 日发布。公告指出，Apple 在底层系统中实施了一系列安全增强措施，旨在提升整体平台的安全性。这些改进不针对特定漏洞，而是通过加固系统组件、改进内存保护机制、增强加密算法实现等方式，减少潜在攻击面。由于公告未提供具体技术细节，尚不清楚这些改进是否涉及已知安全问题或主动防御更新。建议用户保持系统版本最新，以受益于这些安全提升。

该输入为阿里云安全公告，但实际内容仅为公安备案号（浙公网安备 33010602009975号），未包含任何漏洞或安全公告的详细信息，无法提取有效威胁或修复建议。

Pocsuite 是由知道创宇安全研究团队开发并维护的一款开源远程漏洞测试框架。该框架旨在辅助安全研究人员进行 Web 安全漏洞的验证与测试，是知道创宇安全研究能力的基础组件。Pocsuite 支持模块化设计，允许用户自定义 PoC 脚本，从而高效地检测和验证各类安全漏洞。目前，该框架持续更新，社区活跃，已被广泛应用于安全研究、渗透测试等领域。然而，本次输入内容仅为框架介绍，并未涉及具体的漏洞发现、安全公告或修复建议，因此无法提供针对特定漏洞的技术细节或影响评估。

输入内容为知道创宇旗下Seebug漏洞平台的简介。Seebug是一个国内知名的安全漏洞社区，收录了大量漏洞及PoC，旨在为安全研究人员和防御方提供漏洞参考与学习资源。但输入中未包含任何具体的漏洞公告、CVE编号、受影响产品或严重性评估，因此无法生成特定的漏洞摘要。作为防御方，应关注该平台发布的最新漏洞信息，以便及时采取修复措施。

该内容为知道创宇云安全（Yunaq）网站安全防护服务的广告介绍，而非安全漏洞公告。文中声称其拥有十年Web安全防护经验，构建了多层安全防护体系，结合云端大数据与安全CDN加速，可识别并拦截DDoS攻击、DNS攻击、CC攻击等。该服务旨在保护网站安全。但未提供任何具体漏洞信息、CVE编号或技术细节。

该输入并非安全公告或漏洞信息，而是京公网安备备案号的相关记录，来源为360CERT，但内容为空。无法提取任何技术细节或安全风险。

该公告为360CERT发布的网络出版许可证信息，标题为“（总）网出证（京）字第281号”，内容涉及360公司获得的网络出版服务资质，并非安全漏洞或威胁情报。公告中未提及任何CVE、漏洞、攻击或影响，因此不涉及安全风险。

输入内容为京网文〔2020〕6051-1195号，来源360CERT，链接指向360.cn/licence1.html，发布时间为2026年7月5日。该内容并非安全公告，而是一个网络文化经营许可证（ICP备案）编号，无任何漏洞信息、影响产品、修复建议等安全相关内容。因此无法生成安全漏洞摘要。

本次公告未提供具体的漏洞相关信息，内容仅为备案号等非安全内容。

A vulnerability was identified in itsourcecode Online Hotel Management System 1.0. The affected element is an unknown function of the file /admin/login.php. The manipulation of the argument email leads to sql injection. The attack may be initiated remotely. The exploit is publicly available and might be used.

Crypt::DSA versions before 1.22 for Perl draw the DSA signing nonce and private key from a biased random generator, leading to private-key recovery. "Crypt::DSA::Util::makerandom forces the high bit of every value it returns to obtain an exactly N-bit integer for prime search. The signing nonce and the private key are drawn from makerandom. Because the high bit is always set, the result is not un

A vulnerability was determined in 666ghj BettaFish up to 1.2.1. Impacted is the function _deduplicate_results of the file InsightEngine/agent.py of the component InsightEngine search-result Deduplication. Executing a manipulation can lead to partial string comparison. The attack can be launched remotely. The exploit has been publicly disclosed and may be utilized. The pull request to fix this issu

A flaw has been found in SourceCodester Multi-Vendor Online Grocery Management System 1.0. Affected by this vulnerability is the function cancel_order of the file classes/Master.php. Executing a manipulation can lead to improper authorization. The attack may be performed from remote. The exploit has been published and may be used.

A security vulnerability has been detected in SourceCodester Multi-Vendor Online Grocery Management System 1.0. This impacts the function update_settings_info of the file classes/SystemSettings.php of the component Setting Handler. Such manipulation of the argument content[] leads to code injection. The attack can be executed remotely. The exploit has been disclosed publicly and may be used.

A vulnerability was found in HdrHistogram up to 2.2.2. This issue affects the function org.HdrHistogram.DoubleHistogram.recordValue of the file src/main/java/org/HdrHistogram/DoubleHistogram.java of the component Range Check. Performing a manipulation results in incorrect comparison. The attack is only possible with local access. The exploit has been made public and could be used. The project was

A vulnerability has been found in SourceCodester Multi-Vendor Online Grocery Management System 1.0. Affected by this issue is the function cancel_order of the file classes/Master.php of the component POST Parameter Handler. The manipulation of the argument ID leads to sql injection. It is possible to initiate the attack remotely. The exploit has been disclosed to the public and may be used.

A vulnerability was detected in SourceCodester Multi-Vendor Online Grocery Management System 1.0/5.7.26. Affected is the function save_shop_type of the file classes/Master.php of the component POST Parameter Handler. Performing a manipulation results in sql injection. The attack is possible to be carried out remotely. The exploit is now public and may be used.

A vulnerability was found in SourceCodester Multi-Vendor Online Grocery Management System 1.0. This affects the function save_client of the file classes/Users.php of the component Registration Handler. The manipulation of the argument Name results in sql injection. It is possible to launch the attack remotely. The exploit has been made public and could be used.

A weakness has been identified in SourceCodester Multi-Vendor Online Grocery Management System 1.0. This affects the function save_users of the file classes/Users.php. This manipulation causes improper authorization. Remote exploitation of the attack is possible. The exploit has been made available to the public and could be used for attacks.

A security flaw has been discovered in CodeAstro Apartment Visitor Management System 1.0. The impacted element is an unknown function of the file /apartment-visitor/add-apartment.php. The manipulation of the argument apartmentno results in sql injection. The attack may be launched remotely. The exploit has been released to the public and may be used for attacks.

A vulnerability has been found in HdrHistogram up to 2.2.2. This vulnerability affects the function recordValueWithCount of the file src/main/java/org/HdrHistogram/AbstractHistogram.java of the component AbstractHistogram. Such manipulation of the argument Count leads to state issue. The attack can only be performed from a local environment. The exploit has been disclosed to the public and may be

A vulnerability was detected in HdrHistogram up to 2.2.2. Affected by this issue is the function org.HdrHistogram.AbstractHistogram.decodeFromCompressedByteBuffer of the file src/main/java/org/HdrHistogram/AbstractHistogram.java. The manipulation of the argument lengthOfCompressedContents results in uncontrolled memory allocation. The attack needs to be approached locally. The exploit is now publi

A security vulnerability has been detected in code-projects Assessment Management 1.0. This affects an unknown part of the file /admin/remove-user.php. The manipulation of the argument ID leads to cross site scripting. It is possible to initiate the attack remotely. The exploit has been disclosed publicly and may be used.

A vulnerability has been found in itsourcecode Hospital Management System 1.0. Impacted is an unknown function of the file /patientappointment.php. Such manipulation of the argument patiente leads to sql injection. The attack can be launched remotely. The exploit has been disclosed to the public and may be used.

A vulnerability was found in code-projects Online Job Portal 1.0. The affected element is an unknown function of the file login.php. Performing a manipulation of the argument txtUser/txtPass results in sql injection. The attack may be initiated remotely. The exploit has been made public and could be used.

A vulnerability was detected in code-projects Assessment Management 1.0. This vulnerability affects unknown code of the file /lecturer/marking-scheme.php. The manipulation of the argument smarksrange[] results in sql injection. It is possible to launch the attack remotely. The exploit is now public and may be used.

A weakness has been identified in code-projects Assessment Management 1.0. Affected by this issue is some unknown functionality of the file admin/view-users.php. Executing a manipulation of the argument User can lead to cross site scripting. The attack may be performed from remote. The exploit has been made available to the public and could be used for attacks.

A flaw has been found in code-projects Assessment Management 1.0. This issue affects some unknown processing of the file /lecturer/marking-scheme.php of the component Database Query Handler. This manipulation of the argument squestions[] causes sql injection. The attack can be initiated remotely. The exploit has been published and may be used.

A flaw has been found in HdrHistogram up to 2.2.2. This affects the function org.HdrHistogram.AbstractHistogram.decodeFromByteBuffer of the file src/main/java/org/HdrHistogram/AbstractHistogram.java. This manipulation of the argument numberOfSignificantValueDigits causes uncontrolled memory allocation. The attack can only be executed locally. The exploit has been published and may be used. The pro

A vulnerability has been found in connorskees grass up to 0.13.4. The impacted element is the function grass_compiler::selector::extend/grass_compiler::evaluate::visitor. The manipulation leads to denial of service. The attack must be carried out locally. The exploit has been disclosed to the public and may be used. The project maintainer explains: "DoS vulnerabilities are generally fine in Sass c

A vulnerability was determined in SourceCodester Simple and Nice Shopping Cart Script 1.0. This impacts an unknown function of the file /admin/mensproductdeletequery.php. This manipulation of the argument user_id causes sql injection. Remote exploitation of the attack is possible. The exploit has been publicly disclosed and may be utilized.

A vulnerability was found in SourceCodester Simple and Nice Shopping Cart Script 1.0. This affects an unknown function of the file /admin/login.php of the component Admin Login. The manipulation of the argument Username results in sql injection. The attack may be launched remotely. The exploit has been made public and could be used.

A security vulnerability has been detected in code-projects Online Voting System up to 0.x/1.0. This issue affects the function test_input of the file /authentication.php of the component Login. Such manipulation of the argument adminUserName/adminPassword leads to sql injection. It is possible to launch the attack remotely. The exploit has been disclosed publicly and may be used.

A vulnerability was detected in code-projects Online Voting System 1.0. Impacted is the function test_input of the file /saveVote.php. Performing a manipulation of the argument voterName/voterEmail/voterID/selectedCandidate results in sql injection. The attack can be initiated remotely.

A vulnerability was determined in SourceCodester Class and Exam Timetabling System 1.0. Affected by this vulnerability is an unknown functionality of the file /edit_course.php. Executing a manipulation of the argument ID can lead to sql injection. The attack can be executed remotely. The exploit has been publicly disclosed and may be utilized.

A vulnerability was identified in SourceCodester Class and Exam Timetabling System 1.0. Affected by this issue is some unknown functionality of the file /edit_class2.php. The manipulation of the argument ID leads to sql injection. The attack is possible to be carried out remotely. The exploit is publicly available and might be used.

A vulnerability was identified in SourceCodester Simple and Nice Shopping Cart Script 1.0. Affected is an unknown function of the file /admin/girlsproductdeletequery.php. Such manipulation of the argument user_id leads to sql injection. The attack can be executed remotely. The exploit is publicly available and might be used.

A flaw has been found in connorskees grass up to 0.13.4. The affected element is the function grass_compiler::raw_to_parse_error of the component UTF-8 Character Handler. Executing a manipulation can lead to denial of service. The attack is restricted to local execution. The exploit has been published and may be used. In Issue #117 with similar structure the project maintainer explains: "DoS vulne

A vulnerability was found in CodeAstro Apartment Visitor Management System 1.0. Affected is an unknown function of the file /index.php of the component Login. Performing a manipulation of the argument Username results in sql injection. Remote exploitation of the attack is possible. The exploit has been made public and could be used.

A weakness has been identified in onnx up to 1.21.x. This vulnerability affects the function convPoolShapeInference_opset19 of the file onnx/defs/nn/old.cc of the component onnxruntime. This manipulation causes out-of-bounds read. It is possible to initiate the attack remotely. The exploit has been made available to the public and could be used for attacks. Patch name: a7bf3a0f1d18bb62575236ef6e49

A vulnerability has been found in CodeAstro Ecommerce Website 1.0. This impacts an unknown function of the file /ecommerce-website-php/customer/my_account.php?edit_account. Such manipulation of the argument c_name leads to sql injection. The attack may be launched remotely. The exploit has been disclosed to the public and may be used.

A flaw has been found in itsourcecode Hospital Management System 1.0. This affects an unknown function of the file /patient.php. This manipulation of the argument editid causes sql injection. The attack may be initiated remotely. The exploit has been published and may be used.

A security vulnerability has been detected in kirilkirkov Ecommerce-CodeIgniter-Bootstrap up to 13fd582aaf49aeab7438acc0fc3eb973a1f5e6a7. The affected element is the function getCartItems in the library application/libraries/ShoppingCart.php. The manipulation of the argument shopping_cart leads to deserialization. The attack can be initiated remotely. The exploit has been disclosed publicly and ma

A weakness has been identified in kirilkirkov Ecommerce-CodeIgniter-Bootstrap up to 23105f25dadf57b4314fc015a63a7c6e910c89df. Impacted is the function do_upload_others_images of the file application/modules/vendor/controllers/AddProduct.php of the component Vendor Image Manager. Executing a manipulation of the argument folder can lead to path traversal. It is possible to launch the attack remotely

A security flaw has been discovered in kirilkirkov Ecommerce-CodeIgniter-Bootstrap up to 222ff31c06687b1c6d0e1ab63953f82c3674c52b. This issue affects some unknown processing of the file application/modules/vendor/controllers/AddProduct.php of the component Vendor Multi-Image Endpoint. Performing a manipulation of the argument folder results in path traversal. It is possible to initiate the attack

A vulnerability was identified in kirilkirkov Ecommerce-CodeIgniter-Bootstrap up to 213babdbaa949e94557246414db0130e01394517. This vulnerability affects the function checkForPostRequests of the file application/core/MY_Controller.php of the component Subscribed Emails Admin Page. Such manipulation of the argument User-Agent leads to cross site scripting. The attack may be performed from remote. Th

A vulnerability was determined in kirilkirkov Ecommerce-CodeIgniter-Bootstrap up to 49b20f53de2b7ec34e920b11c863f1491d911a04. This affects an unknown part of the file /index.php/api/product/set of the component Hidden REST API Endpoint. This manipulation of the argument title/description causes cross site scripting. The attack is possible to be carried out remotely. The exploit has been publicly d

A vulnerability was found in kirilkirkov Ecommerce-CodeIgniter-Bootstrap up to 95dfa8cebbb87ab46ae450643a07241274a74dce. Affected by this issue is the function setReferrer of the file application/core/MY_Controller.php of the component Trusted Backend Interface. The manipulation of the argument href results in open redirect. The attack can be executed remotely. The exploit has been made public and

Trail of Bits fickling versions up to and including 0.1.10 do not include the Python standard library modules _posixsubprocess, site, and atexit in the UNSAFE_IMPORTS denylist (fickle.py). Because these modules are absent from the denylist, fickling's check_safety() function returns LIKELY_SAFE with zero findings for pickle payloads that invoke dangerous functions including _posixsubprocess.fork_e

In Trail of Bits fickling versions up to and including 0.1.11, the UnsafeImportsML analysis pass unconditionally calls AnalysisContext.shorten_code(node) on every import node it inspects, regardless of whether the import is flagged as unsafe. This call registers the shortened code representation in the shared AnalysisContext.reported_shortened_code set. When the MLAllowlist analysis pass subsequen

A flaw has been found in RT-Thread up to 5.2.2. Affected is the function read/write/sys_ioctl of the file components/lwp/lwp_syscall.c of the component Parameter Handler. Executing a manipulation can lead to divide by zero. The attack may be launched remotely. The exploit has been published and may be used. The pull request to fix this issue awaits acceptance.

In multi-tenanted deployments, the application consent management mechanism fails to correctly isolate consent scopes between tenants. Consent granted by a user for a specific SaaS application within one tenant can be incorrectly applied to SaaS applications with the same name in other tenants, leading to unintended cross-tenant consent sharing. This vulnerability may result in the exposure of us

In the Linux kernel, the following vulnerability has been resolved: KVM: x86: Fix shadow paging use-after-free due to unexpected role Commit 0cb2af2ea66ad ("KVM: x86: Fix shadow paging use-after-free due to unexpected GFN") fixed a shadow paging mismatch between stored and computed GFNs; the bug could be triggered by changing a PDE mapping from outside the guest, and then deleting a memslot. Th

HestiaCP panel cronjob feature is affected by a broken access control vulnerability. Low privilege users can modify the panel cronjob to execute scripts HestiaCP management scripts with passwordless sudo. This could result in the takeover of administrator users in the application and the underlying webserver.

A weakness has been identified in NousResearch hermes-agent up to 2026.4.30. The impacted element is the function AIAgent.run_conversation of the file run_agent.py of the component HTTP API. This manipulation of the argument todos causes denial of service. The attack can be initiated remotely. The exploit has been made available to the public and could be used for attacks. The vendor was contacted

A vulnerability was identified in omec-project amf up to 2.0.2/2.1.1. Impacted is an unknown function of the file /go/src/amf/ngap/handler.go of the component NGSetupRequest Handler. The manipulation leads to denial of service. It is possible to initiate the attack remotely. The exploit is publicly available and might be used. The identifier of the patch is 34bc6724acc97dba1f8691e586da95b042cb612d

In the Linux kernel, the following vulnerability has been resolved: af_unix: Set gc_in_progress to true in unix_gc(). Igor Ushakov reported that unix_gc() could run with gc_in_progress being false if the work is scheduled while running: Thread 1 Thread 2 Thread 3 -------- -------- -------- unix_schedule_gc()

A vulnerability was determined in omec-project amf up to 2.1.1. This issue affects the function RRCInactiveTransitionReport of the component NGAP Message Handler. Executing a manipulation can lead to denial of service. The attack may be performed from remote. The exploit has been publicly disclosed and may be utilized. This patch is called 34bc6724acc97dba1f8691e586da95b042cb612d. A patch should b

In the Linux kernel, the following vulnerability has been resolved: KVM: SEV: Require in-GHCB scratch area if GHCB v2+ is in use As per the GHCB spec, when using GHCB v2+ require the software scratch area to reside in the GHCB's shared buffer. Note, things like Page State Change (PSC) requests _rely_ on this behavior, as the guest can't provide a length when making the request, i.e. the size of

In the Linux kernel, the following vulnerability has been resolved: ipv6: account for fraggap on the paged allocation path In __ip6_append_data(), when the paged-allocation branch is taken (MSG_MORE / NETIF_F_SG / large fraglen), alloclen and pagedlen are computed as alloclen = fragheaderlen + transhdrlen; pagedlen = datalen - transhdrlen; datalen already includes fraggap (datalen = length +

myVesta is affected by an authenticated remote code execution vulnerability. Low privileged users can insert arbitrary commands as a part of the v_ftp_user parameter when deleting FTP usernames. This could result in the execution of commands as the admin user or takevoer of the admin user in myVesta.

A flaw has been found in itsourcecode Hospital Management System 1.0. Affected by this issue is some unknown functionality of the file /medicine.php. This manipulation of the argument editid causes sql injection. Remote exploitation of the attack is possible. The exploit has been published and may be used.

A vulnerability was found in jairiidriss restaurant-website-php-mysql up to 521428b5b612449df0cf4a5d15ee40cba67f3d35. This vulnerability affects unknown code of the file /admin/ajax_files of the component AJAX Endpoint. Performing a manipulation results in missing authentication. The attack is possible to be carried out remotely. The exploit has been made public and could be used. This product ado

PHPIPAM is affected by an authenticated local file inclusion vulnerability that allows users with access to the API to execute/include arbitrary PHP files on the web server's file system. The API is not enabled by default on installations.

A vulnerability was detected in Open5GS up to 2.7.7. Affected by this vulnerability is the function amf_nnrf_handle_nf_discover of the file src/amf/nnrf-handler.c of the component AMF. The manipulation results in denial of service. The attack may be launched remotely. The exploit is now public and may be used. The patch is identified as fb5f67703de0213fb9c6e6ef3b48b6c1707e9503. It is best practice

In nltk/nltk versions 3.9.3 and earlier, five Stanford interface classes (StanfordPOSTagger, StanfordNERTagger, StanfordParser, StanfordDependencyParser, and StanfordNeuralDependencyParser) are vulnerable to untrusted JAR code execution. These classes accept user-controllable JAR paths and execute them via the `java()` function, which invokes `subprocess.Popen()` without integrity verification. Th

picklescan before 0.0.34 fails to detect the _operator.methodcaller built-in function when scanning pickle files for malicious code. Attackers can craft malicious pickle payloads using _operator.methodcaller that evade detection and execute arbitrary code when loaded by pickle.load().

picklescan before 0.0.30 fails to detect malicious pickle files that invoke torch.utils.bottleneck.__main__.run_autograd_prof function. Attackers can embed undetected code in pickle files that executes during deserialization, enabling remote code execution.

picklescan before 0.0.34 fails to detect _operator.attrgetter function calls in pickle payloads, allowing attackers to bypass security checks. Remote attackers can craft malicious pickle files using _operator.attrgetter in reduce methods to execute arbitrary code when pickle.load() processes the file.

picklescan before 0.0.28 fails to detect malicious torch.utils.bottleneck.__main__.run_cprofile function calls in pickle files, allowing attackers to bypass safety checks. Remote attackers can embed undetected code in pickle files to achieve arbitrary code execution when victims load the files.

picklescan before 0.0.28 fails to detect malicious pickle files that use torch.utils.data.datapipes.utils.decoder.basichandlers in reduce methods, allowing attackers to bypass safety checks. Remote attackers can embed undetected malicious code in pickle files that executes during deserialization, enabling remote code execution.