markitdown

Pass

Audited by Gen Agent Trust Hub on Mar 4, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it extracts and processes text from external, potentially untrusted files.
  • Ingestion points: Document contents from PDF, DOCX, PPTX, XLSX, HTML, images, audio, and ZIP archives (SKILL.md).
  • Boundary markers: Absent. No delimiters or 'ignore' instructions are used to separate converted content from the agent's system prompt.
  • Capability inventory: Shell execution of the markitdown CLI and basic text processing tools (wc, grep).
  • Sanitization: Absent. Extracted content is passed to stdout without validation or escaping.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 4, 2026, 12:16 AM