elevenlabs-tts
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 external installation instructions hosted on GitHub for the
beltCLI utility from theinference-shrepository.- [COMMAND_EXECUTION]: The skill utilizes thebeltcommand-line tool to execute remote applications, manage user authentication, and perform audio synthesis.- [INDIRECT_PROMPT_INJECTION]: The skill presents an injection surface by processing user-supplied text for synthesis. - Ingestion points: The
textfield within the JSON input for thebelt app runcommand. - Boundary markers: No explicit boundary markers or instruction-ignore wrappers are used for the user input.
- Capability inventory: The skill is restricted to using the
belttool via Bash. - Sanitization: No sanitization or filtering of the input text is described.
Audit Metadata