extract-transcript

Pass

Audited by Gen Agent Trust Hub on Apr 28, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted content from YouTube transcripts. Malicious actors could include instructions within a video's captions to subvert the agent's behavior when the transcript is analyzed.
  • Ingestion points: The extract_transcript.py script fetches data from an external, third-party source (YouTube) using the youtube-transcript-api library.
  • Boundary markers: There are no delimiters or protective instructions in the output to help the agent distinguish between the transcript data and valid commands.
  • Capability inventory: The script can write data to any file path provided as the output_file argument, and it performs network requests to access YouTube's transcript services.
  • Sanitization: The transcript text is not sanitized, filtered, or validated before being returned to the agent or written to the local filesystem.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 28, 2026, 04:20 PM