claude-design-system-architect

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to fetch and analyze content from arbitrary websites to establish design brand inputs. This ingestion of untrusted data, combined with the agent's ability to write configuration files and execute shell commands, creates a surface for indirect prompt injection attacks where a malicious website could attempt to influence the agent's actions.
  • Ingestion points: Step 1 involves using WebFetch and WebSearch to analyze existing websites or reference materials provided by the user or found online.
  • Boundary markers: The instructions do not specify any delimiters or safety markers to prevent the agent from obeying instructions that might be hidden within the fetched web content.
  • Capability inventory: The skill allows access to Write, Bash, WebSearch, and WebFetch tools, allowing the agent to create files (tailwind.config, tokens.css) based on the analyzed data.
  • Sanitization: There is no requirement in the skill instructions to sanitize, filter, or validate the content retrieved from external URLs before it is used in the generation process.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 07:25 AM
Security Audit — agent-trust-hub — claude-design-system-architect