geofeed-tuner
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill downloads IP prefix data from remote URLs specified by the user to perform analysis and tuning.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external CSV files; it implements structural validation and HTML escaping during report generation to mitigate potential injection risks from external data.
- [DYNAMIC_EXECUTION]: The skill generates and executes Python scripts to perform complex data processing tasks, including validation, API lookups, and HTML template rendering. This is a primary function of the skill and is confined to a dedicated working directory.
- [REMOTE_CODE_EXECUTION]: The skill uses CDNs for Leaflet, H3-JS, and Bootstrap Icons to render maps and icons in the generated HTML report. These are established well-known services.
Audit Metadata