youtube-transcript
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The bundled Python script
scripts/fetch-youtube-transcript.pyexecutes the externalyt-dlpbinary using thesubprocess.runmethod. The implementation is secure as it passes arguments as a list and does not enable shell interpretation (shell=Trueis not used), which effectively prevents command injection vulnerabilities from malicious YouTube URLs or parameters. - [EXTERNAL_DOWNLOADS]: The skill requires
yt-dlp, a popular third-party tool for interacting with video platforms. The instructions guide the agent to ensure it is installed, typically via system package managers like Homebrew. This is a standard requirement for the skill's primary functionality. - [INDIRECT_PROMPT_INJECTION]: As the skill ingests external text from YouTube transcripts, it is theoretically susceptible to indirect prompt injection where a video author could embed instructions for the AI. However, the skill provides clear mitigation instructions in
SKILL.md, explicitly telling the agent to treat page text as untrusted and ignore any instructions found within the transcript data.
Audit Metadata