fal-music
Pass
Audited by Gen Agent Trust Hub on Mar 25, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill interacts with the FAL.ai API via queue.fal.run to perform text-to-audio generation. All network communication is directed to well-known service endpoints associated with the skill's primary function.- [CREDENTIALS_UNSAFE]: The skill handles API keys using standard practices. It supports environment variables (FAL_KEY) and local JSON storage. In the
cmd_configfunction, the script explicitly applieschmod 600to the configuration file, ensuring that sensitive credentials are only readable by the owner.- [DATA_EXFILTRATION]: The code includes a lookup for a sibling directory (../fal-video-skill/config.json) to share an API key between related tools. This cross-skill access is documented and serves a functional purpose for a suite of AI tools from the same developer without exfiltrating data to unknown third parties.- [COMMAND_EXECUTION]: The tool performs standard file system operations and network requests using therequestslibrary. These actions are scoped to the intended use case of downloading and saving generated audio files to a localoutputdirectory.- [PROMPT_INJECTION]: The skill processes user-supplied text prompts for music generation. To mitigate indirect risks, it implements asanitize_filenamefunction that uses regex to strip potentially dangerous characters from the prompt before using it as a filename for generated content.
Audit Metadata