The Agent Skills Directory

[SAFE]: The skill provides robust guidelines for secure React and Next.js development, including specific warnings about multi-tenant cache isolation and server action authorization.
[SAFE]: No malicious patterns, such as prompt injection, obfuscation, or unauthorized remote code execution, were detected. All external links point to official documentation of reputable technology providers (Vercel, Tailwind Labs, Testing Library).
[SAFE]: Recommended dependencies are standard industry libraries and tools used for their intended purpose in frontend development and testing.
[INDIRECT_PROMPT_INJECTION]: The skill includes functionality for reviewing external codebase files and locale strings, which presents a theoretical ingestion surface for untrusted data.
Ingestion points: Processes .tsx, .jsx, .ts, and JSON locale files through its provided instructions.
Boundary markers: No specific delimiters are defined in the instructions for separating untrusted code content from agent instructions.
Capability inventory: The skill provides instructions for code generation, review, state management, and locale synchronization.
Sanitization: The skill recommends the use of Zod for validating inputs in server actions and provide guidance on avoiding unsafe use of dangerouslySetInnerHTML for untrusted data.

react-frontend