The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides templates for functions that convert text to speech. If these functions are used to process untrusted external data in a production environment, they represent an injection surface where maliciously crafted text could attempt to influence the audio generation output.
Ingestion points: The text property in the TTSOptions interface within src/elevenlabs/tts-service.ts and the text parameter in the generate_speech function within elevenlabs_service.py.
Boundary markers: None are included in the code snippets to distinguish between instructions and data.
Capability inventory: The patterns include capabilities to write generated audio data to the local file system using fs.createWriteStream (TypeScript) and open(output_path, 'wb') (Python).
Sanitization: No text sanitization or validation logic is included in the provided code templates.

elevenlabs-sdk-patterns