The Agent Skills Directory

[SAFE]: No security threats were identified. The skill correctly identifies itself as a tool for improving frontend interfaces and provides legitimate resources to that end.\n- [INDIRECT_PROMPT_INJECTION]: The skill utilizes a Python script (scripts/design_audit.py) to scan local project files for design-related patterns. This constitutes an indirect prompt injection surface as it processes untrusted file content. However, the ingestion is limited to specific UI-related file extensions, ignores common binary/meta directories, and the resulting data is used for structural analysis rather than executable command generation.\n- [COMMAND_EXECUTION]: The skill workflow instructs the agent to execute standard repository commands such as npm run build or npm test. While this involves executing shell commands, they are project-specific verification steps performed at the user's implicit direction within the context of a development workflow.

design-engineer