claude-design
Pass
Audited by Gen Agent Trust Hub on Sep 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow that processes untrusted external data which could contain embedded instructions.
- Ingestion points: In the 'Design Principle' and 'Reading Documents and Assets' sections, the agent is directed to read various file types (Markdown, HTML, CSS, JS, TS, etc.) from the local filesystem or remote GitHub repositories.
- Boundary markers: The skill lacks instructions for using delimiters or specific warnings to ignore natural language instructions found within the analyzed files.
- Capability inventory: The agent is empowered to write multiple HTML, CSS, and JS files to the disk and is encouraged to verify them using local environment tools.
- Sanitization: There are no requirements for the agent to sanitize, escape, or validate the content extracted from external source files before it influences the generation of design artifacts.
Audit Metadata