Grav < 1.0.0-rc.16 Authentication Bypass via token URL Parameter
CVE-2026-62386
The Grav API plugin (getgrav/grav-plugin-api) before 1.0.0-rc.16 accepts JWT access tokens through the ?token= URL query
2026-07-18 归档
今日共收录 1491 条安全情报，包含 418 个 CVE，581 篇安全通告，以及 194 篇研究论文。
← 返回归档列表OpenClaw 2026.4.14 < 2026.5.26 SSRF via Browser Snapshot
CVE-2026-62227
OpenClaw 2026.4.14 before 2026.5.26 contain a server-side request forgery vulnerability in browser snapshot routes that
OpenClaw 2026.3.28 < 2026.5.19 Authorization Bypass via Browser Act Route
CVE-2026-62226
OpenClaw 2026.3.28 before 2026.5.19 contain an authorization bypass vulnerability in the browser act route that fails to
wger: Privilege escalation via trainer-login session chaining allows gym trainers to impersonate gym managers
CVE-2026-43978
wger is a free, open-source workout and fitness manager. In versions prior to 2.6, a gym trainer can escalate their sess
Kirby: Content locks disclose IDs and emails of inaccessible users from `users.access/list` permissions
CVE-2026-45334
Kirby is an open-source content management system. In versions prior to 4.9.1 and 5.4.1, the content-locking feature ret
Kirby: Cross-site scripting (XSS) from list field content in the site frontend
CVE-2026-44175
Kirby is an open-source content management system. In versions prior to 4.9.1 and 5.4.1, Kirby did not securely sanitize
PyMdown Extensions: Regression in pymdownx.snippets reintroduces sibling-prefix path traversal bypass despite restrict_base_path
CVE-2026-46338
PyMdown Extensions is a set of extensions for the Python-Markdown markdown project. From 10.0.1 until 10.21.3, pymdownx.
Sensitive content disclosure via OpenTelemetry spans in AgentCore Python SDK
CVE-2026-15737
AWS Bedrock AgentCore Python SDK is an open-source Python library that provides client tools for building AI agents on t
Jupyter Enterprise Gateway launches remote Jupyter Notebook kernels across distributed clusters like Apache Spark, Kubernetes, and Docker Swarm. Versions 2.0.0rc1 and above prio...
CVE-2026-44180
Jupyter Enterprise Gateway launches remote Jupyter Notebook kernels across distributed clusters like Apache Spark, Kuber
WireGuard Easy through 15.3.0, fixed in commit 66b292b, contains a cryptographically weak one-time link token generation vulnerability that allows unauthenticated network attack...
CVE-2026-63089
WireGuard Easy through 15.3.0, fixed in commit 66b292b, contains a cryptographically weak one-time link token generation
Authentication bypass using an alternate path or channel in Microsoft Edge (Chromium-based)...
CVE-2026-57980
Authentication bypass using an alternate path or channel in Microsoft Edge (Chromium-based) allows an unauthorized attac
An issue in Generic OEM UZ801_v2.1 4G LTE Router V3.4.3 allows a remote attacker to execute...
CVE-2026-52199
An issue in Generic OEM UZ801_v2.1 4G LTE Router V3.4.3 allows a remote attacker to execute arbitrary code via the sbin/
AWS-JDBC Wrapper: Privilege Escalation in Aurora PostgreSQL instance
CVE-2026-11400
Aurora PostgreSQL is a fully managed relational database engine that's compatible with PostgreSQL. The team has identif
vLLM: Speech-to-text upload size limit is enforced after full UploadFile read
CVE-2026-55646
## Summary Current-head vLLM documents `VLLM_MAX_AUDIO_CLIP_FILESIZE_MB` as the maximum audio file size accepted by the
Jazzer: Coverage-Guided Fuzzing for Semantic Vulnerabilities in the Java Ecosystem.
本文介绍了Jazzer，一种针对Java生态系统的覆盖率引导的模糊测试工具，专门用于发现语义漏洞。与传统的仅关注内存安全错误的模糊测试不同，Jazzer通过结合覆盖率反馈和语义分析，能够检测更高级别的逻辑缺陷，例如不正确的状态转换、业务逻辑
PickleFuzzer: A Case Study in Fuzzing for Discrepancies Between Python Pickle Implementations.
Python原生的序列化协议pickle虽然功能强大，但存在已知的安全风险，尤其是在传输不可信数据或保存机器学习模型时。为了缓解风险，开发者通常会在反序列化时限制导入的模块，或使用静态和动态分析工具检测恶意负载。然而，这些方法依赖于对Pic
TDXRay: Microarchitectural Side-Channel Analysis of Intel TDX for Real-World Workloads.
该论文针对英特尔TDX（Trust Domain Extensions）这一基于虚拟机的机密计算环境，系统研究了其微架构侧信道攻击面。TDX旨在保护租户代码和数据免受特权云操作系统的侵害，但往往将微架构侧信道排除在威胁模型之外。作者发现了四
Investigating the Impact of Dark Patterns on LLM-Based Web Agents.
随着用户越来越多地依赖基于大型语言模型（LLM）的Web代理来自动化在线任务，这些代理可能会遇到暗模式（dark patterns）——即欺骗性的用户界面设计，旨在操纵用户做出非本意的决策。尽管暗模式主要针对人类用户，但它们对基于LLM的通
MoPE: A Mixture of Password Experts for Improving Password Guessing.
本文针对密码猜测任务中现有数据驱动模型对密码结构差异忽略导致训练偏差的问题，提出了一种混合密码专家框架（MoPE）。研究背景指出，当前密码强度评估方法通常将密码视为统一数据，忽视了不同密码的结构模式（如固定长度数字串、字母数字混合等），导致
Toward Efficient Membership Inference Attacks Against Federated Large Language Models: A Projection Residual Approach.
本文针对联邦大语言模型（FedLLMs）中的成员推断攻击（MIA）问题展开研究。FedLLMs允许多方协作微调大语言模型而无需共享原始数据，从而解决资源有限和隐私保护等挑战。然而，即使数据不离本地，共享的梯度仍然可能通过成员推断攻击泄露敏感