whichllm before 0.5.16 contains a code injection vulnerability in the run and snippet commands that allows a remote attacker who controls a HuggingFace repository to achieve arbitrary code execution by crafting a malicious GGUF filename containing double quotes or other special characters. The script generation function in cli.py interpolates HuggingFace-derived values, including GGUF variant filenames from the Hub API siblings rfilename field, directly into Python source code without escaping, allowing the crafted filename to break out of the generated string literal and execute injected code on the user's machine before any model download occurs.

whichllm versions before 0.5.16 contain a code injection vulnerability in the run and snippet commands. A remote attacker who controls a HuggingFace repository can achieve arbitrary code execution by crafting a malicious GGUF filename containing double quotes or special characters. The vulnerable script generation function in cli.py interpolates HuggingFace-derived values, including GGUF variant filenames from the Hub API siblings rfilename field, directly into Python source code without proper escaping. This allows the crafted filename to break out of the generated string literal and execute injected code on the victim's machine. Exploitation occurs before any model download takes place, lowering the bar for attack. The vulnerability has been patched in version 0.5.16.