elevenlabs-voice-isolator
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references installation guidelines from the inference-sh GitHub repository.
- [COMMAND_EXECUTION]: The skill executes the
beltCLI tool for audio processing. This execution is restricted to the specific command prefixbeltwithin theBashtool, minimizing the risk of arbitrary command injection. - [INDIRECT_PROMPT_INJECTION]: The skill exposes an attack surface by processing external audio data.
- Ingestion points: Audio file URLs are ingested via the
audiofield in the command's JSON input. - Boundary markers: The use of JSON formatted strings provides a boundary between data and instructions.
- Capability inventory: The skill uses the
beltCLI to process audio files and generate output. - Sanitization: No specific input validation or sanitization of the provided URLs is documented in the skill instructions.
Audit Metadata