The Agent Skills Directory

[COMMAND_EXECUTION]: The skill extensively uses the rawgenai CLI tool to perform text-to-speech, multi-speaker dialogue generation, and voice management operations across various providers.
[EXTERNAL_DOWNLOADS]: The instructions require installing the rawgenai utility from the author's Homebrew tap (brew install WHQ25/tap/rawgenai). This is documented as a vendor-provided tool.
[PROMPT_INJECTION]: The skill processes untrusted text from users or files to generate audio, creating an inherent surface for indirect prompt injection where instructions embedded in the data can manipulate the speech model's output style or behavior.
Ingestion points: Text passed to rawgenai tts via command arguments, the --file flag, or stdin.
Boundary markers: Absent; the skill does not use delimiters or instructions to ignore embedded commands in the processed text.
Capability inventory: The skill executes shell commands that interact with multiple third-party API providers (OpenAI, Google, ElevenLabs, DashScope, MiniMax, Kling, Seed).
Sanitization: No sanitization or validation of input text is specified in the skill's instructions.

agent-speak