skills/mathews-tom/armory/to-markdown/Gen Agent Trust Hub

to-markdown

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill requires installing several packages from the Python Package Index (PyPI), including markitdown, trafilatura, and playwright, which are well-known libraries for document and web processing.
  • [COMMAND_EXECUTION]: Shell commands are used for environment setup, including uv pip install for Python packages and playwright install to fetch the Chromium browser. The skill also mentions using Node.js tools via npx.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external files and URLs, converting them into text for LLM consumption.
  • Ingestion points: Files (PDF, Word, etc.) and remote URLs processed via markitdown and custom fetch strategies.
  • Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions for the resulting text.
  • Capability inventory: The toolset includes file system access (writing markdown), network fetching, and shell command execution.
  • Sanitization: The conversion to Markdown strips non-textual formatting but does not verify the semantic content of the extracted text for malicious instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 09:09 AM