The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes system commands (ffmpeg, ffprobe) to extract audio and subtitles. The Python script correctly uses argument lists for subprocess calls to prevent shell injection.- [EXTERNAL_DOWNLOADS]: The instructions recommend installing well-known libraries (openai, python-dotenv) from the official Python Package Index (PyPI).- [DATA_EXFILTRATION]: Audio data is transmitted to OpenAI or Azure APIs for transcription. This is the documented purpose of the skill and utilizes credentials stored in a local environment file (~/.transcribe_video.env) provided by the user.- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection if the processed video content or embedded subtitles contain instructions designed to manipulate the agent during the results reporting phase.\n
Ingestion points: Video file metadata, embedded subtitle streams, and API transcription responses.\n
Boundary markers: Absent; the transcript preview is displayed directly to the agent without specific delimiters or isolation.\n
Capability inventory: Shell command execution via ffmpeg and ffprobe, file system write access for transcripts, and network access to external APIs.\n
Sanitization: No sanitization is performed on the extracted text before it is presented to the agent.

transcribe-video