The Agent Skills Directory

[SAFE]: The skill interacts exclusively with authorized PostPlus Cloud endpoints for media processing and transcription. Communication is secured via session tokens managed through standard local configuration files.
[COMMAND_EXECUTION]: A shared library script includes functionality to execute yt-dlp via Python for media acquisition. This is a well-known utility used appropriately within the context of media production tasks.
[EXTERNAL_DOWNLOADS]: The skill downloads transcription results and artifacts from the vendor's hosted environment. These operations are restricted to the primary function of the skill and use structured request envelopes to maintain security boundaries.
[PROMPT_INJECTION]: The skill includes an optional prompt parameter in the transcription request. This is a standard feature of Whisper models used to guide transcription accuracy and does not pose a threat to the agent's system instructions.

audio-transcription