design
Pass
Audited by Gen Agent Trust Hub on May 9, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXFILTRATION]: The skill extracts information from the local codebase, including component source code, Tailwind classes, and project configuration (e.g.,
.subframe/sync.json), to provide context to Subframe's remote MCP tools likedesign_page. This data transmission is necessary for the tool to generate designs that match the user's existing project. - [PROMPT_INJECTION]: The skill ingests untrusted data from the local environment (the user's code) and passes it to external tools without explicit sanitization or boundary markers. This represents an indirect prompt injection surface where instructions hidden in code could influence the design generation process, though this is a functional characteristic of context-aware design tools.
Audit Metadata