youtube-transcript

Pass

Audited by Gen Agent Trust Hub on Aug 5, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses yt-dlp to extract video metadata and subtitles. It also includes the yt-dlp -U command, which allows the utility to self-update its binary.
  • [EXTERNAL_DOWNLOADS]: Fetches transcript data from https://deepapi.co and interacts with the internet to update the local yt-dlp scraper tool.
  • [PROMPT_INJECTION]: The skill processes external data (YouTube transcripts), creating a surface for indirect prompt injection if the agent interprets the saved text as instructions in a future context.
  • Ingestion points: External YouTube transcript data fetched via DeepAPI or local scraping (SKILL.md).
  • Boundary markers: Absent; the skill does not wrap the output in delimiters or provide instructions for the agent to treat the resulting transcript as untrusted data.
  • Capability inventory: The skill has access to curl for network requests, yt-dlp for scraping, and python3 for text transformation.
  • Sanitization: Basic text cleaning, HTML unescaping, and whitespace normalization are performed via a Python script, but there is no filtering for malicious prompt instructions embedded in the transcripts.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 5, 2026, 08:17 AM
Security Audit — agent-trust-hub — youtube-transcript