url-to-markdown

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions for installing various third-party processing tools from reputable sources.
  • Fetches go-readability from github.com/go-shiori/go-readability using go install.
  • Installs the markitdown package from Microsoft's official repository via pip.
  • Installs standard utility packages readability-lxml and trafilatura via pip.
  • [PROMPT_INJECTION]: The skill is designed to ingest content from arbitrary external URLs, creating a surface for indirect prompt injection.
  • Ingestion points: External URLs provided to tools for content extraction (SKILL.md).
  • Boundary markers: Not explicitly specified in the skill instructions.
  • Capability inventory: Uses shell-based CLI tools (curl, trafilatura, markitdown) to fetch and process data.
  • Sanitization: Leverages content extraction algorithms (Mozilla Readability, Trafilatura) which naturally filter out much of the HTML structure but may not prevent embedded textual instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 10:33 PM
Security Audit — agent-trust-hub — url-to-markdown