document-extractor

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the markitdown package and its optional feature groups (e.g., pdf, docx, all) from PyPI using uv or pipx. It also references documentation from Microsoft's official GitHub repository.
  • [COMMAND_EXECUTION]: The skill executes the markitdown CLI tool to convert various file formats into Markdown. This includes command-line operations for file conversion, plugin listing, and integration with Azure Document Intelligence.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data from external files (PDF, Word, Excel, etc.) which may contain hidden instructions intended to manipulate the agent's behavior after the document is converted to text.
  • Ingestion points: SKILL.md and references/feature-groups.md (via markitdown <input_file> commands)
  • Boundary markers: Absent; the skill does not explicitly instruct the agent to ignore instructions within the converted content.
  • Capability inventory: File system read/write access and network access for Azure Document Intelligence services.
  • Sanitization: Absent; the skill relies on the underlying tool for conversion without additional filtering of the output.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 09:32 PM