Lucene search
+L

3 matches found

Kitploit
Kitploit
added 2026/09/01 7:37 p.m.5 views

CVE-2025-51859

CVE-2025-51859 漏洞描述 Chaindesk 是一个用于构建 AI 智能体的 Web 应用程序,其智能体聊天组件存在存储型跨站脚本(XSS)漏洞。攻击者可以通过构造 AI 智能体并使其系统提示词指示底层大语言模型(LLM)在聊天响应中嵌入恶意脚本载荷(例如基于 SVG 的 XSS),从而在客户端执行任意脚本。当用户与这类恶意智能体交互,或访问包含 XSS 载荷的会话直接链接时,脚本会在用户浏览器中执行。成功利用该漏洞可导致敏感信息(如 JWT 会话令牌)被窃取,进而可能造成账户劫持。会话 URL...

6.5CVSS6.2AI score0.00435EPSS
SaveExploits1
CVE
CVE
added 2026/07/22 11:21 a.m.82 views

CVE-2026-65015

CVE-2026-65015 affects n8n and is exploitable in versions before 2.30.1 through the AI Agents feature. The issue arises because the node-execution tool does not perform proper authorization checks, enabling a Project Viewer user to escalate privileges by chatting with an agent that has node tools...

8.8CVSS5.5AI score0.00324EPSS
SaveExploits0References2Affected Software1
Github Security Blog
Github Security Blog
added 2026/06/18 1:57 p.m.32 views

PraisonAI AgentTeam.launch exposes unauthenticated remote agent listing and invocation endpoints

PraisonAI AgentTeam.launch exposes unauthenticated remote agent invocation endpoints Summary PraisonAI's documented Python AgentTeam.launch / Agents.launch HTTP server starts externally reachable agent invocation endpoints without any authentication enforcement. The current implementation registe...

5.7AI score
SaveExploits0References2Affected Software1
Rows per page
Query Builder