elevenlabs-stt
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 and tools. Specifically, it suggests installing the
beltCLI tool vianpx skills add belt-sh/cliand links to a setup guide hosted on GitHub atraw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md.- [COMMAND_EXECUTION]: The skill uses theBash(belt *)tool to execute various commands such asbelt loginandbelt app run. These commands are used to interact with the ElevenLabs STT and forced-alignment applications by passing JSON input.- [INDIRECT_PROMPT_INJECTION]: This skill has an attack surface for indirect prompt injection as it processes external audio files and text data. - Ingestion points: The skill accepts a URL to an audio file (
audio) and a string of text (text) to be aligned. - Boundary markers: The data is encapsulated within a JSON object passed to the
--inputargument of thebeltcommand. - Capability inventory: The skill utilizes shell command execution via the
beltCLI to process the data. - Sanitization: The instructions do not specify any sanitization or validation of the input content before it is processed by the external service.
Audit Metadata