baoyu-url-to-markdown

Pass

Audited by Gen Agent Trust Hub on Mar 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: Spawns a local browser instance (Google Chrome, Microsoft Edge, or Chromium) using child_process.spawn. The skill attempts to locate these executables in standard system paths or respects a URL_CHROME_PATH environment variable. The skill also manages a persistent browser profile directory in the user's home folder to maintain login sessions and state.
  • [PROMPT_INJECTION]: This skill has an attack surface for indirect prompt injection.
  • Ingestion points: Fetches content from arbitrary URLs in scripts/main.ts.
  • Boundary markers: Output consists of YAML metadata followed by content; no specific delimiters or warnings are used to isolate untrusted web content from the agent's context.
  • Capability inventory: Includes process spawning (scripts/cdp.ts), file system writes (scripts/main.ts), and network requests (scripts/media-localizer.ts).
  • Sanitization: Uses jsdom and defuddle to parse HTML, which may strip active scripts but preserves text-based instructions that can influence the agent.
  • [EXTERNAL_DOWNLOADS]: Fetches and saves media assets (images, videos) from arbitrary URLs to the local filesystem when the media localization feature is active.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 14, 2026, 06:32 AM