The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill connects to ElevenLabs' official API (api.elevenlabs.io) to perform text-to-speech generation. ElevenLabs is a well-known service, and the network communication is restricted to its documented endpoints.
[CREDENTIALS_UNSAFE]: The skill implements secure secret management by explicitly instructing the user to store the ELEVENLABS_API_KEY in environment variables or a .env file, avoiding the risk of hardcoded credentials in the source code.
[INDIRECT_PROMPT_INJECTION]: The skill processes text input from external files, command-line arguments, or standard input. While this constitutes an attack surface, the data is properly serialized into a JSON payload for the ElevenLabs API, and the script uses sanitization (slugifying) for generated filenames to prevent path traversal issues.

elevenlabs-tts