autoshorts
Pass
Audited by Gen Agent Trust Hub on Jul 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The CLI tool invokes
ffmpegandffprobeto perform frame-accurate video cutting and dimension analysis. These subprocesses are executed using structured argument lists, which avoids the security risks associated with shell command injection. - [EXTERNAL_DOWNLOADS]: The skill downloads AI models for the Whisper library to enable local video transcription. This is a standard and expected operation for the skill's primary function.
- [DATA_EXFILTRATION]: Video content and metadata are transmitted to
Google Geminifor multimodal analysis and toUpload-Postfor social media scheduling. These network operations target legitimate service providers and are essential for the skill's stated workflow. - [PROMPT_INJECTION]: The skill processes video transcripts as part of its AI-driven analysis. The implementation uses clear boundary markers and structured prompts to manage external inputs and ensure the agent remains within its operational scope.
Audit Metadata