The Agent Skills Directory

[SAFE]: No security threats or malicious patterns were identified during the analysis.
Prompt Injection: No attempts to override agent instructions or bypass safety filters were detected. The skill uses standard instructional language to guide UI generation.
Data Exposure & Exfiltration: No hardcoded credentials, sensitive file path access, or unauthorized network operations were found. Network references (e.g., in CSS URLs or package mentions) are purely for well-known services or illustrative purposes.
Obfuscation: No Base64-encoded commands, zero-width characters, homoglyphs, or other hidden content techniques were identified.
Unverifiable Dependencies & RCE: The skill provides code examples and mentions standard, well-known packages (like flutter_staggered_grid_view or @react-native-community/blur) as recommendations for developers, but does not execute any remote code or install unverifiable packages at runtime.
Indirect Prompt Injection: While the skill processes user requests for UI styles, it does not possess exploitable capabilities (file writes, subprocess execution, or network sends) that could be abused via indirect injection.
Dynamic Context Injection: The main SKILL.md file does not contain any !command syntax that would execute shell commands at load time.

design-it