speech-to-text
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions for installing the official SDKs for the service, including the 'elevenlabs' and 'python-dotenv' packages for Python, and '@elevenlabs/elevenlabs-js', '@elevenlabs/client', and '@elevenlabs/react' for Node.js.
- [COMMAND_EXECUTION]: Contains standard setup and utility commands using 'pip', 'npm', and 'curl' to interact with the ElevenLabs API.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external audio and video content to produce text transcripts. This introduces a surface for indirect prompt injection if the agent processes the resulting text without treating it as untrusted data.
- Ingestion points: The transcribed text returned by the 'speech_to_text.convert' method and real-time streaming events ('partial_transcript', 'committed_transcript') across all implementation examples.
- Boundary markers: None are present in the provided code snippets to distinguish the transcribed content from system instructions.
- Capability inventory: The skill uses network tools to communicate with 'api.elevenlabs.io' and standard file system tools to read local audio files.
- Sanitization: The examples do not include any sanitization or validation of the transcribed text output.
Audit Metadata