POC详情: de6595faecb8ec7ecfb23e4633a6b91845279c5d

来源
关联漏洞
标题: BentoML 代码问题漏洞 (CVE-2025-32375)
描述:BentoML是BentoML开源的一个开源模型服务库。用于使用 Python 构建高性能和可扩展的人工智能应用程序。 BentoML 1.4.8之前版本存在代码问题漏洞,该漏洞源于不安全的反序列化,可能导致执行任意代码。
描述
This repository includes everything needed to run a PoC exploit for CVE-2025-32375 in a Docker environment. It runs the latest vulnerable version of BentoML (1.4.7).
介绍
# Setup for the vulnerable version.

Start the vulnerable container:
```sh
docker compose up
```

Create a listener (e.g. ncat):
```sh
ncat -klnv 1337
```

Run the exploit:
```sh
python3 exploit.py
```

You should receive an HTTP request in the ncat window with the contents of the OS command (`id` by default) if everything worked as expected.

# Credits
+ Vulnerability found by [SeaW1nd](https://twitter.com/SeaW1nd1405)
+ Based on the testing setup by [VickyTheViking](https://github.com/VickyTheViking)
文件快照

[4.0K] /data/pocs/de6595faecb8ec7ecfb23e4633a6b91845279c5d ├── [ 109] docker-compose.yml ├── [ 408] Dockerfile ├── [ 534] exploit.py ├── [ 508] README.md └── [4.0K] workdir ├── [ 172] bentofile.yaml ├── [ 246] create.py └── [ 460] service.py 1 directory, 7 files
神龙机器人已为您缓存
备注
    1. 建议优先通过来源进行访问。
    2. 如果因为来源失效或无法访问,请发送邮箱到 f.jinxu#gmail.com 索取本地快照(把 # 换成 @)。
    3. 神龙已为您对POC代码进行快照,为了长期维护,请考虑为本地POC付费,感谢您的支持。