skills/tt-a1i/hermes-agent/design-md/Gen Agent Trust Hub

design-md

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill uses npx to download and execute the @google/design.md package from the npm registry. This is the official toolset for managing the DESIGN.md specification.
  • [COMMAND_EXECUTION]: The skill executes shell commands via npx to perform design token validation (linting), version diffing, and exporting to formats like Tailwind CSS or W3C DTCG JSON. These commands operate on local DESIGN.md files.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied design specifications, which constitutes a potential ingestion point for untrusted data. However, it mitigates risk through the following:
  • Ingestion points: Reads existing DESIGN.md content provided by the user or found in the project directory.
  • Boundary markers: None explicitly defined in the prompt instructions, though the skill organizes data into a formal YAML/Markdown structure.
  • Capability inventory: Employs write_file for file creation and npx for executing the validation CLI.
  • Sanitization: Relies on the @google/design.md lint command to structurally validate tokens and markdown headings before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 12:48 PM
Security Audit — agent-trust-hub — design-md