design-sound
Pass
Audited by Gen Agent Trust Hub on Nov 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill performs audio processing and interaction with external APIs (ElevenLabs, Fal.ai) as described in its purpose. No malicious behavior or security risks were identified during the analysis.- [COMMAND_EXECUTION]: Multiple scripts, including
mix.py,fit_music.py,fetch_web.py,video_foley.py,fal_foley.py, andab_test.py, utilizesubprocess.runto call external utilities likeffmpeg,ffprobe,yt-dlp, andnvidia-smi. These calls are implemented using list-based arguments and do not invoke a shell, which mitigates command injection risks.- [EXTERNAL_DOWNLOADS]: The skill includes functionality to fetch audio content from reputable and well-known sources, such asarchive.org,incompetech.com,opengameart.org, andyoutube.com. These operations are consistent with the skill's role as a sound design tool.- [DYNAMIC_EXECUTION]:score.pyemploysimportlibto load a local project script (bank_samples.py), andvideo_foley.pydynamically modifiessys.pathto interface with a local model installation. These are standard Python practices for managing local project dependencies.- [OBFUSCATION]: The automated detection ofeval()insound_library.pywas determined to be a false positive. The code calls the PyTorch.eval()method to set an AI model to evaluation mode, which is unrelated to the dangerous Pythoneval()built-in function.
Audit Metadata