space-video-subtitle
Pass
Audited by Gen Agent Trust Hub on Aug 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides documentation for standard
ffmpegcommands used to encapsulate or burn subtitles into video files (SKILL.md). - [PROMPT_INJECTION]: The skill is designed to process untrusted data (transcriptions from audio/video files) through an AI model for proofreading. This creates a surface for Indirect Prompt Injection.
- Ingestion points: Transcription text generated by ASR backends like SenseVoice or Whisper (SKILL.md).
- Boundary markers: None mentioned for the AI proofreading instructions.
- Capability inventory: Command execution capabilities via
ffmpeg(SKILL.md). - Sanitization: The instructions focus on correcting terminology and grammar without mentioning data sanitization.
- Assessment: This behavior is central to the skill's primary purpose and no malicious injection patterns were observed in the provided instructions.
Audit Metadata