lov-migrate-camera-media
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes system binaries including
diskutil,ioreg, andsyncto identify storage volumes and verify hardware connection parameters. It additionally invokesffmpegandffprobeto decode frames and verify the integrity of media files. - [INDIRECT_PROMPT_INJECTION]: The skill processes external media files and XML metadata, presenting a potential ingestion surface for untrusted data. Ingestion points: Files are read from the source volume in
scripts/camera_media.pyandscripts/check_media.py. Boundary markers: User authorization is required for cleanup, and path validation is performed. Capability inventory: Includes file read/write/delete operations and system command execution. Sanitization: Implements directory traversal checks (safe_rel) and SHA-256 hash verification for all transferred files.
Audit Metadata