The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill documentation includes instructions to install the Whisper library from OpenAI's official GitHub repository. Because OpenAI is a trusted organization, this external download is considered safe.
[COMMAND_EXECUTION]: The skill utilizes the whisper command-line tool and the Python subprocess module to handle audio file transcription. These operations are essential for the skill's functionality and are performed locally.
[PROMPT_INJECTION]: The skill was reviewed for indirect prompt injection surfaces (Category 8). 1. Ingestion points: Audio files provided for transcription. 2. Boundary markers: None mentioned for the transcription output. 3. Capability inventory: Subprocess calls and file system access for processing. 4. Sanitization: Not explicitly implemented in the provided patterns. While the skill processes untrusted audio, no malicious code or instructions were detected within the skill's own implementation.

openai-whisper