The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a bundled Python script scripts/text_to_speech.py to interact with the OpenAI API. It includes logic to handle batch jobs via JSONL files and manage file system paths for output storage.
[EXTERNAL_DOWNLOADS]: The skill requires the installation of the openai Python package from the official Python Package Index (PyPI). It directs the agent to use uv or pip for installation.
[DATA_EXPOSURE]: The skill requires an OPENAI_API_KEY to function. It explicitly instructs the agent never to ask the user to paste the key into the chat, instead advising them to set it as a local environment variable.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-provided text and style instructions which are then sent to the OpenAI Audio API.
Ingestion points: User input provided via CLI flags (--input, --instructions) or read from local files (--input-file, --instructions-file) in scripts/text_to_speech.py.
Boundary markers: Absent. The skill does not use delimiters or specific instructions to prevent the model from following commands embedded in the input text.
Capability inventory: File system writes (audio output and temporary JSONL files), and outbound network access to the OpenAI API.
Sanitization: Absent. Input text is passed directly to the API after basic validation of character length.

speech