Gen Agent Trust Hub

This skill provides a tool for executing Python code in a sandboxed remote environment via the inference.sh platform. It supports a wide range of pre-installed libraries for data science, web scraping, and media processing. While it allows for arbitrary code execution, this is the core intended functionality of the skill. A minor risk of indirect prompt injection exists if the executed code is generated from untrusted sources.

The provider classified the analyzed snapshot as NONE risk. This result applies to the content available at audit time.