The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill takes natural language text as input to generate audio via ElevenLabs. This represents an indirect prompt injection surface where untrusted data could influence audio output.
Ingestion points: Text inputs in client.text_to_speech.convert, and prompt strings in client.text_to_sound_effects.convert and client.music.compose.
Boundary markers: None identified; instructions are directly interpolated into API calls.
Capability inventory: The skill has the capability to write files to the local file system (e.g., voiceover.mp3, music.mp3) and make outbound network requests to the ElevenLabs API.
Sanitization: No specific sanitization or filtering of input text is described.
[COMMAND_EXECUTION]: The skill documents and utilizes command-line tools for media processing and project management.
Evidence: Examples include uv run tools/voiceover.py and ffmpeg for audio conversion and manifest generation.
Context: These commands are standard for the skill's primary purpose of media generation and are intended to be executed within a development environment.

elevenlabs