design-md

Warn

Audited by Gen Agent Trust Hub on Oct 11, 2026

Risk Level: MEDIUMREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill executes the getdesign package from the npm registry using npx getdesign@latest. This downloads and executes code from an external source at runtime. It also executes @google/design.md from Google's official repository via npx.
  • [EXTERNAL_DOWNLOADS]: The skill fetches design system reference files (DESIGN.md) from 68 external brands using the getdesign CLI. The fetch_all_design_md.sh script automates the download of these files into a local library.
  • [COMMAND_EXECUTION]: The skill relies on shell command execution via a bash script (fetch_all_design_md.sh) to manage its local library and environment. It uses npx within these scripts to interact with remote packages.
  • [INDIRECT_PROMPT_INJECTION]: The skill acts as a data ingestion point for remote content that may influence downstream agent behavior.
  • Ingestion points: Fetches remote markdown files via npx getdesign and scans local CSS files in generate_design_md.py (SKILL.md).
  • Boundary markers: The instructions do not specify the use of delimiters or safety warnings for the fetched remote design documents.
  • Capability inventory: The skill environment has the capability to execute shell commands, run npx packages, and perform file system writes.
  • Sanitization: There is no documented sanitization or filtering of the fetched markdown content to prevent embedded instructions from affecting the agent.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Oct 11, 2026, 05:16 PM