ml-feature-engineering
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE]: The skill consists entirely of static Markdown files and a CSV template. There are no executable scripts (.py, .js, .sh), binaries, or configuration files that trigger command execution.
- [INDIRECT_PROMPT_INJECTION]: The skill outlines a process for gathering external data such as 'candidate feature hypotheses' and 'data sources'. While these inputs represent a potential injection surface, the skill has no associated tools or scripts that could execute instructions found within that data. The risk is mitigated by the lack of agent capabilities in this specific skill.
Audit Metadata