product-design-get-context

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to perform local shell operations. In references/local-prototype-preflight.md, it directs the agent to run a bootstrap script (node bootstrap-prototype.mjs) and execute npm install. Additionally, SKILL.md instructs the agent to run a 'preflight script' from a referenced user-context skill when local shell access is available.
  • [EXTERNAL_DOWNLOADS]: The instruction to run npm install in references/local-prototype-preflight.md initiates the download of packages from the public npm registry to set up the local development environment.
  • [PROMPT_INJECTION]: The skill includes directive language such as 'Critical Overrides' and 'Hard boundary' in SKILL.md and references/critical-overrides.md. These instructions are designed to override default assistant behavior and enforce a strict design-brief gate before any UI generation or implementation occurs.
  • [DATA_EXPOSURE]: The skill encourages the agent to read local project information, such as codebase paths, browser preferences, and design tokens, to provide context-aware design suggestions. While typical for design assistants, this involves accessing project metadata and configuration files.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 06:05 PM
Security Audit — agent-trust-hub — product-design-get-context