design-md

Pass

Audited by Gen Agent Trust Hub on Aug 18, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill analyzes untrusted repository content, such as existing design documents, code comments, and configuration files, to generate the root DESIGN.md. This represents a potential attack surface where malicious instructions embedded in a project's CSS or documentation could influence the agent's output.
  • Ingestion points: package.json, globals.css, tailwind.config.*, component files, and existing DESIGN.md or brand guidelines.
  • Boundary markers: The instructions require the agent to "cite real files" and trace every claim back to evidence, which provides structural grounding, though no explicit delimiters are used to wrap ingested data.
  • Capability inventory: The skill utilizes file-writing capabilities to create DESIGN.md, .agents/skills/frontend-design/SKILL.md, and update AGENTS.md.
  • Sanitization: No explicit sanitization of ingested code or documentation content is defined before the agent processes it.
  • [DYNAMIC_EXECUTION]: The skill generates a new agent skill file at .agents/skills/frontend-design/SKILL.md. This behavior is the intended primary purpose of the skill to distribute portable craft doctrine. The content is generated from a static, local template (references/skill-frontend-design.md) rather than remote sources or untrusted inputs.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 18, 2026, 08:24 AM
Security Audit — agent-trust-hub — design-md