read-youtube-transcript
Pass
Audited by Gen Agent Trust Hub on Jul 11, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The script
scripts/get_transcript.pyincludes logic to dynamically install theyoutube-transcript-apipackage usingpipif it is not detected in the environment. This is a standard and well-known library for fetching YouTube subtitles. - [COMMAND_EXECUTION]: The skill uses
subprocess.check_callto invoke the local Python interpreter for installing missing dependencies. This usage is restricted to setup operations required for the skill's primary functionality. - [PROMPT_INJECTION]: The skill processes transcript content from YouTube videos, which represents an indirect prompt injection surface. External data from video subtitles is untrusted and could contain embedded instructions intended to influence the agent's behavior.
- Ingestion points:
scripts/get_transcript.py(via YouTube API). - Boundary markers: None; the transcript text is handled as raw data for output or translation.
- Capability inventory: The skill can execute shell commands for setup and perform local file system writes via the
-ooutput flag. - Sanitization: No filtering or validation is performed on the transcript text before it is returned to the agent context.
Audit Metadata