audiocraft-audio-generation
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references the official
audiocraftlibrary and related tools, instructing users to install them from PyPI or official repositories under the Facebook Research GitHub organization. - [COMMAND_EXECUTION]: Provides standard shell commands for environment setup, including Python package installation via
pip, cloning repositories withgit, and installing system-level dependencies like FFmpeg. - [INDIRECT_PROMPT_INJECTION]: The skill documents a workflow where user-provided text descriptions are processed by audio generation models.
- Ingestion points: Untrusted text enters the system via the
descriptionslist in local Python scripts and thepromptfield in the FastAPI and Gradio interface examples (e.g.,SKILL.md). - Boundary markers: No specific delimiters or instructions for the model to ignore embedded commands are included in the prompt construction.
- Capability inventory: The provided code samples perform file system writes (saving .wav files) and network operations (automatic model weight downloads triggered by the library).
- Sanitization: No explicit input validation or sanitization of the user prompts is shown in the instructional examples.
Audit Metadata