The Agent Skills Directory

[DATA_EXFILTRATION]: The skill performs network requests to Google's official Gemini API endpoint (generativelanguage.googleapis.com) to generate audio from text. It securely manages authentication by reading API keys from environment variables rather than hardcoding them.
[INDIRECT_PROMPT_INJECTION]: The skill processes external JSON data to define speech generation turns, which creates a surface for potential indirect prompt injection attacks where malicious data could influence the agent's or the TTS model's behavior.
Ingestion points: scripts/gemini_tts.py reads data from files provided via the --input command-line argument.
Boundary markers: The script implements strict schema validation, checking for allowed keys (model, voice, speakers, turns) and rejecting unknown fields.
Capability inventory: The skill has the ability to make outbound network requests and write WAV files to the local file system.
Sanitization: The script validates data types and ensures input is correctly JSON-encoded before it is sent to the API endpoint.

gemini-tts