You need to agree to share your Hugging Face information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this model content.

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

Malicious Pickle Proof-of-Concept (PoC)

This repository contains a deliberately crafted .pkl file that demonstrates a well-known security vulnerability in Python's pickle module β€” arbitrary code execution via unsafe deserialization.

πŸ” Vulnerability Summary

  • Format: .pkl
  • Type: Insecure Deserialization β†’ Arbitrary Code Execution (ACE)
  • Method: Uses __reduce__() in a custom Python class to trigger execution of an OS command at load time.
  • PoC File: malicious.pkl

⚠️ Security Warning

This file is intentionally malicious. Do not deserialize (pickle.load) it on any machine you don't control. It will run a harmless OS command for demonstration, but the technique can be used to run any code.

πŸ§ͺ Reproduction Instructions

Run the following to reproduce the issue safely:

python3 -c "import pickle; pickle.load(open('malicious.pkl', 'rb'))"

Expected output: Malicious pickle loaded successfully! UID=...

The command is executed via os.system() during deserialization.

πŸ“‰ Impact Remote Code Execution on the system loading the pickle file.

Full compromise if attacker supplies the pickle.

Demonstrates the risk of loading untrusted .pkl files.

βœ… Mitigations Do not use pickle.load() on untrusted input.

Prefer secure serialization formats such as safetensors, JSON, or ONNX.

If unavoidable, use strong sandboxing or input validation.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support