The Agent Skills Directory

[COMMAND_EXECUTION]: Executes curl to send audio data to OpenAI's transcription endpoint. All variables in the shell script are properly quoted to prevent command injection.\n- [DATA_EXFILTRATION]: Sends user-specified audio files to api.openai.com for processing. As this is a well-known service and the primary purpose of the skill, it is considered safe.\n- [CREDENTIALS_UNSAFE]: Uses the OPENAI_API_KEY environment variable for authentication. The documentation correctly advises users to manage this secret through environment variables or local configuration files rather than hardcoding it.

openai-whisper-api