markitdown

Pass

Audited by Gen Agent Trust Hub on Jul 14, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Installs the markitdown and markitdown-mcp packages from the standard Python package registry. These are tools developed by Microsoft for extracting content from various document formats.
  • [COMMAND_EXECUTION]: The skill executes the markitdown utility via shell commands to process local files. Installation is performed using the --user flag to ensure execution within user space without requiring elevated permissions.
  • [PROMPT_INJECTION]: The skill has an indirect prompt injection surface because it processes untrusted external data (PDF, Office, etc.) and converts it into a format intended for LLM consumption.
  • Ingestion points: Files processed by the markitdown command in SKILL.md.
  • Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded commands within the converted text.
  • Capability inventory: The skill has access to Bash, Read, and Write tools.
  • Sanitization: Relies on the library's internal logic to convert structured data into Markdown text.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 14, 2026, 08:28 AM
Security Audit — agent-trust-hub — markitdown