Tag
OpenAI 的红队演练发现 Brotli 压缩库存在未认证远程代码执行漏洞，影响 Node.js、Bun、Envoy 等广泛部署的软件；相关修复已通过 CVE 和下游安全更新发布。
Two critical PHP zero-day vulnerabilities (CVE-2026-109618 and CVE-2026-109619) affecting PHP 5.4 and later are being exploited in the wild, prompting an emergency PHP release and outages across countless sites, with OpenAI and Wiz involved in the research and disclosure.
A heap use-after-free in PHP's session.upload_progress handler (CVE-2026-109618 / -109619) allows attackers to hijack freed bucket pointers and achieve remote code execution via zend_call_function; the writeup includes verified PoC chains for PHP 8.4.4 and 8.2.34 along with pre-patch mitigations.
多个 Plex Media Server 漏洞（Plex2Shell，CVE-2026-96652 等）正被野外利用以投放加密货币挖矿程序，研究人员通过蜜罐捕获了实际攻击流量与 IOC。
GitLab警示一个最大严重级别的CI/CD正则表达式编译器整数溢出漏洞（CVE-2026-97229），可让远程未认证攻击者在服务器上执行任意代码，官方已发布补丁并敦促自托管用户立即升级。
研究者分享了在 Google 开发者文档站点发现的 9 万美元远程代码执行漏洞：Flask 会话 cookie 中 pickle 序列化的 OAuth flow 对象泄露了 client_secret，而该 secret 同时又是会话签名密钥，攻击者可伪造 cookie 并通过 pickle.loads 实现 RCE。
Attackers have been exploiting a critical unauthenticated command injection vulnerability (CVE-2026-73570) in the Zimbra Collaboration Suite to steal email backups and credentials, with over 274 instances compromised despite a July patch. Microsoft disclosed active exploitation including web shells, reverse shells, and mailbox data exfiltration across multiple industries.
A remote code execution vulnerability in LuaRocks.org was exploited between July and August 2026, leading to exposed user data and credentials. Users are advised to take immediate security actions such as changing passwords and revoking keys.
This article reports on a critical security vulnerability in WordPress that allows unauthenticated path traversal leading to conditional remote code execution, with setup instructions from the development repository.
HEIF Heist is a class of security vulnerabilities in image parsers like libheif that allow remote code execution and data exposure through crafted HEIF, HEIC, or AVIF images, affecting major services such as OpenAI, Slack, and GitHub Enterprise.
A pre-authentication remote code execution vulnerability (CVE-2026-32746) has been discovered in GNU inetutils Telnetd, impacting multiple operating systems due to its legacy code base.
Security researchers discovered critical vulnerabilities in the default GitHub Actions configurations for Anthropic's Claude Code, Google's Gemini CLI, and OpenAI's Codex, allowing remote code execution through unauthenticated issues. A related privilege escalation flaw in Google's ADK repo highlights systemic weaknesses in the CI/CD scaffolding for these AI agents.
Calif Research demos WeWorm, a zero-click worm spreading through WeChat calls, with AI enabling rapid exploit development.
A remote code execution vulnerability has been found in Unitree robots, and it is wormable, enabling attackers to infect and control multiple robots.
This CVE describes a prototype pollution vulnerability in n8n that can be exploited for remote code execution through a Git node gadget chain.
Researchers used publicly available AI models to discover a critical zero-click, memory-corruption vulnerability in Zoom's annotation feature, allowing remote code execution on all meeting participants across all platforms. Zoom has issued fixes, but the research highlights how AI is lowering the barrier to nation-state-level exploits.
This article breaks down the MCP security attack chain behind CVE-2025-53773, where a single prompt injection lets a developer agent modify its own configuration and achieve remote code execution. It details the hop-by-hop escalation and the key control that stops it.
A critical remote code execution vulnerability (CVE-2026-66066) has been discovered in Ruby on Rails' Active Storage when using the default Vips image processor, affecting Rails 7.x and 8.x default configurations. Patches have been released and immediate upgrading is recommended.
Research reveals a critical vulnerability in Azure Cosmos DB's Gremlin API that could have allowed attackers to compromise all databases in the service, including Microsoft's internal ones. Microsoft has fully remediated the issue and no customer action is required.
A heap buffer overflow vulnerability in Dnsmasq (CVE-2026-2291) allows remote code execution via a malicious upstream DNS server. The issue was introduced in version 2.73 and fixed in 2.92rel2 and 2.93.