analyze-ui-video
Pass
Audited by Gen Agent Trust Hub on Jul 31, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The script
scripts/video_frames.pyexecutesffmpegandffprobeusing thesubprocessmodule. These operations are restricted to specific video processing tasks. The script uses argument lists instead of shell strings to prevent command injection, and it validates or sanitizes input parameters like crop dimensions and timestamps.- [DATA_EXFILTRATION]: The skill is designed for local-first analysis. Instructions explicitly tell the agent to keep recordings and frames local. There are no network calls or attempts to transmit data to external servers.- [PROMPT_INJECTION]: The instructions establish a rigorous 'narrate before explaining' protocol that requires the agent to document visible timestamps and facts before performing interpretation, which serves as a defensive reasoning layer against potential adversarial content in video media.
Audit Metadata