video-lens-gallery
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill uses a bash discovery loop to locate its own scripts across several common AI agent configuration directories (e.g.,
~/.agents,~/.claude,~/.cursor). This is a standard mechanism for local skill execution and is not considered malicious. - [COMMAND_EXECUTION]: Executes local Python scripts (
backfill_meta.py,build_index.py) and a shell script (serve_report.sh) to process local files and start a web server. These operations are restricted to the skill's own logic and the user's local directory. - [EXTERNAL_DOWNLOADS]: The gallery interface (
index.html) loads standard assets from well-known services, including web fonts from Google Fonts and video thumbnails from YouTube's image servers.
Audit Metadata