CowAgent 访问控制错误漏洞

CowAgent is an intelligent assistant and scalable agent framework developed by zhayujie’s individual developer. Versions of CowAgent 2.0.4 and earlier contained a security vulnerability related to access control. This vulnerability stemmed from the absence of authentication in the Agent Mode...

MetaGPT 代码注入漏洞

MetaGPT is a multi-agent framework developed by MetaGPT Inc. Versions of MetaGPT 0.8.1 and earlier contained a code injection vulnerability. This vulnerability stemmed from the generatethoughts function in the Tree-of-Thought Solver component’s metagpt/strategy/tot.py file, which could lead to...

PT-2026-32562

Name of the Vulnerable Software and Affected Versions jq versions prior to commit 6374ae0bcdfe33a18eb0ae6db28493b1f34a0a5b Description CLI input parsing allows validation bypass via embedded NUL bytes when reading JSON from files or stdin. The software uses strlen to determine buffer length inste...

MetaGPT 安全漏洞

MetaGPT is a multi-agent framework developed by MetaGPT Inc. Versions of MetaGPT 0.8.1 and earlier contained security vulnerabilities. These vulnerabilities stemmed from the evaluateCode function in the Mineflayer HTTP API component’s file metagpt/environment/minecraft/mineflayer/index.js, which...

XDG Desktop Portal 安全漏洞

XDG Desktop Portal is a frontend service for the desktop application sandbox environment developed by Flatpak. Versions of XDG Desktop Portal prior to 1.20.4 and 1.21.1 contained security vulnerabilities. These vulnerabilities stemmed from the possibility for any Flatpak application to manipulate...

CVE-2026-40159

PraisonAI is a multi-agent teams system. Prior to 4.5.128, PraisonAI’s MCP Model Context Protocol integration allows spawning background servers via stdio using user-supplied command strings e.g., MCP"npx -y @smithery/cli ...". These commands are executed through Python’s subprocess module. By...

LoLLMs 安全漏洞

LoLLMs is a large language and multimodal system personally developed by Saifeddine ALOUI. Versions of LoLLMs prior to 2.2.0 contained a security vulnerability. This vulnerability stemmed from the createpost function not properly cleaning the content provided by users, which could lead to...

S3CDM: A Secret-Sharing-Scheme-Based Cyberattack Detection Model and Its Simulation Implementation

We design and develop a secret-sharing-scheme-based cyberattack detection modelS3CDMthat can detect unauthorized or illegal activities especially insider attacks and protect sensitive information within complex network infrastructures of large organizations. The model splits a secret among a grou...

The vulnerability of the Omnibox address bar in Google Chrome allows attackers to perform spoofing attacks.

The vulnerability of the Omnibox address bar in Google Chrome is related to information representation errors in the user interface. Exploiting this vulnerability allows a malicious actor to carry out spam attacks using a specially created HTML page...

musl libc 安全漏洞

musl libc is an open-source C language standard library developed by musl. It is primarily used in embedded systems and mobile devices. Versions of musl libc up to 1.2.6 contained security vulnerabilities, which were caused by inefficient algorithms and could lead to local attacks...

The vulnerability of the Permissions control element in the Google Chrome browser allows attackers to perform spoofing attacks.

The vulnerability of the Permissions control element in Google Chrome is related to information representation errors in the user interface. Exploiting this vulnerability can allow a malicious actor to carry out phishing attacks using a specially created HTML page...

D-Link DIR-645 安全漏洞

The D-Link DIR-645 is a wireless router produced by D-Link Corporation. Versions 1.01, 1.02, and 1.03 of the D-Link DIR-645 contain security vulnerabilities. These vulnerabilities stem from a stack-based buffer overflow vulnerability in the function hedwigcgimain located in the /cgi-bin/hedwig.cg...

Follow My Eyes: Backdoor Attacks on VLM-Based Scanpath Prediction

Scanpath prediction models forecast the sequence and timing of human fixations during visual search, driving foveated rendering and attention-based interaction in mobile systems where their integrity is a first-class security concern. We present the first study of backdoor attacks against VLM-bas...

Tenda AC15 安全漏洞

The Tenda AC15 is a wireless router produced by the Chinese company Tenda. Version 15.03.05.18 of the Tenda AC15 contains a security vulnerability. This vulnerability arises from incorrect operations with parameters oldPwd/newPwd/cfmPwd in the function websGetVar within the...

Mattermost Plugins 安全漏洞

Mattermost Plugins is a plugin provided by the American company Mattermost, offering powerful feature extensions and tight integration with servers and web/dashboard applications. Versions of Mattermost Plugins 2.1.3.0 and earlier contain security vulnerabilities. These vulnerabilities stem from ...

PraisonAI 安全漏洞

PraisonAI is a low-code multi-agent collaboration framework developed by Mervin Praison. Versions of PraisonAI prior to 4.5.128 contained security vulnerabilities. These vulnerabilities stemmed from treating remotely obtained template files as trusted executable code without performing integrity...

Securing Retrieval-Augmented Generation: A Taxonomy of Attacks, Defenses, and Future Directions

Retrieval-augmented generation RAG significantly enhances large language models LLMs but introduces novel security risks through external knowledge access. While existing studies cover various RAG vulnerabilities, they often conflate inherent LLM risks with those specifically introduced by RAG. I...

OpenClaw 安全漏洞

OpenClaw is an open-source intelligent artificial assistant developed by OpenClaw. Versions of OpenClaw prior to 2026.3.25 contained a security vulnerability. This vulnerability stemmed from the lack of rate limiting in Telegram Webhook authentication, which could lead to brute-force attacks...

Storm-1175 Deploys Medusa Ransomware Within 24 Hours of Flaw Disclosure

Microsoft researchers have uncovered a fast-moving group, Storm-1175, launching high-speed Medusa ransomware attacks against healthcare and education sectors in the UK, US, and Australia by exploiting security flaws in as little as 24 hours...

Insufficient Verification of Data Authenticity

Overview wwbn/avideo is an Audio and Video Platform or simply "A Video Platform". Affected versions of this package are vulnerable to Insufficient Verification of Data Authenticity through the ipn.php process. An attacker can repeatedly increase their wallet balance and renew subscriptions by...