The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill utilizes official SDKs for ElevenLabs, which is a recognized and well-known service in the AI ecosystem.
Evidence: References to elevenlabs (Python) and @elevenlabs/elevenlabs-js (Node.js) in references/installation.md.
[DATA_EXFILTRATION]: All network requests are directed to ElevenLabs' official API infrastructure.
Evidence: The skill documentation specifies communication with https://api.elevenlabs.io/v1/music.
[PROMPT_INJECTION]: The skill processes user-supplied text for music generation, which represents a surface for indirect prompt injection.
Ingestion points: The prompt parameter used in all music.compose methods across SKILL.md and references/api_reference.md.
Boundary markers: The skill does not implement local prompt delimiters, but it relies on the provider's API for safety enforcement.
Capability inventory: The skill performs API calls and local file writes; it does not request shell access, sensitive file reads, or system-level privileges.
Sanitization: Content filtering and copyright protection are managed by the ElevenLabs API, which identifies and rejects unsafe prompts with bad_prompt errors.
[SAFE]: The skill provides secure guidance for API key management by prioritizing environment variables.
Evidence: references/installation.md explicitly recommends storing the ELEVENLABS_API_KEY in environment variables rather than hardcoding it.

music