The Agent Skills Directory

[SAFE]: The skill provides documentation and implementation examples for the official OpenAI Whisper speech recognition model, utilizing trusted dependencies and standard installation procedures.
[COMMAND_EXECUTION]: Instructions include standard system commands for installing the ffmpeg dependency using package managers like apt (with sudo), brew, and choco to satisfy software requirements.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates the processing of external audio data for transcription purposes.
Ingestion points: The model.transcribe() function in SKILL.md and references/languages.md ingests external audio files.
Boundary markers: The initial_prompt parameter is documented for providing context, which can assist in guiding model output and resolving technical terminology.
Capability inventory: The skill uses standard file system access and subprocess execution via the whisper and ffmpeg CLIs for audio processing tasks.
Sanitization: Standard transcription procedures are utilized as expected for automated speech recognition tasks.

whisper