promo-video

Fail

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION] (SAFE): The file scripts/generate_voiceover.py uses subprocess.run to interface with ffmpeg and ffprobe. Analysis shows that all command arguments are passed as lists and filenames are programmatically generated (e.g., section_00.mp3), which prevents command injection vulnerabilities.
  • [EXTERNAL_DOWNLOADS] (SAFE): The script downloads audio data from api.elevenlabs.io. While flagged by automated scanners, this is a legitimate requirement for a voiceover skill. Per the [TRUST-SCOPE-RULE], ElevenLabs is considered a reputable service for this context.
  • [REMOTE_CODE_EXECUTION] (SAFE): The script includes an optional dependency on openai-whisper, which downloads AI models during runtime. Since these are sourced from OpenAI, they are considered trusted for the intended transcription verification purpose.
  • [DATA_EXFILTRATION] (SAFE): The script accesses ELEVEN_LABS_API_KEY from environment variables. This key is only sent to the official ElevenLabs API endpoint to authenticate requests, which is the expected behavior.
Recommendations
  • HIGH: Downloads and executes remote code from: unknown (check file) - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 17, 2026, 06:20 PM