skills/jgraph/drawio-mcp/drawio/Gen Agent Trust Hub

drawio

Pass

Audited by Gen Agent Trust Hub on Sep 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to use the drawio command-line tool for converting Mermaid diagrams to XML and exporting them to various image formats. It also uses platform-specific commands like open, xdg-open, and cmd.exe /c start to display the final output to the user.
  • [EXTERNAL_DOWNLOADS]: The skill fetches Mermaid syntax and XML reference guidelines from the vendor's official GitHub repository (jgraph/drawio-mcp). These are used as non-executable documentation to guide the agent in authoring valid diagram code.
  • [INDIRECT_PROMPT_INJECTION]: As the skill processes user-supplied descriptions to generate diagram content that is eventually passed to a CLI tool, there is a theoretical surface for indirect injection. However, the skill provides specific instructions for handling inputs, such as using node -e with arguments instead of direct shell interpolation, which mitigates standard command injection risks.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 23, 2026, 12:00 PM