youtube-video-transcript
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill retrieves transcript data from an external API endpoint (https://ai-factory.frevana.com/service/youtube-video-transcript), which serves as an ingestion point for untrusted content.
- Ingestion points: Data is fetched from the Frevana API into a temporary file in scripts/get_youtube_video_transcript.sh.
- Boundary markers: There are no explicit instructions or delimiters to prevent the agent from interpreting instructions that might be embedded in the transcript text.
- Capability inventory: The skill can write files to the local filesystem using the --output flag.
- Sanitization: The script validates the JSON structure but does not sanitize or filter the content of the transcript.
- [COMMAND_EXECUTION]: The script scripts/get_youtube_video_transcript.sh executes shell commands including curl, bash, and python3 to perform its tasks.
- [EXTERNAL_DOWNLOADS]: The skill makes network requests to https://ai-factory.frevana.com to retrieve YouTube transcript information.
- [DYNAMIC_EXECUTION]: The bash script uses python3 to execute embedded Python code at runtime for argument parsing and JSON validation. This is a local execution of a static template.
Audit Metadata