kosmos-scientist-guide
Pass
Audited by Gen Agent Trust Hub on Apr 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides structured guidelines for conducting scientific research. It encourages standard practices like setting random seeds, logging hyperparameters, and reporting negative results.
- [COMMAND_EXECUTION]: The skill is designed to guide an agent in writing and executing Python code for machine learning experiments. This is the intended primary purpose of the research framework.
- [EXTERNAL_DOWNLOADS]: References external resources such as the OpenAlex API for literature searches and GitHub for project source. These are used for data retrieval in a research context and do not involve untrusted code execution.
Audit Metadata