The Agent Skills Directory

[SAFE]: The skill consists of markdown documentation and code examples that follow industry-standard performance best practices for React applications. No security vulnerabilities were identified in the provided instructions or scripts.
[EXTERNAL_DOWNLOADS]: The skill references several well-known and trusted external sources, including React.dev, MDN Web Docs, the Vercel blog, and the official lru-cache GitHub repository. These resources are used appropriately for technical context and documentation.
[INDIRECT_PROMPT_INJECTION]: The skill provides instructions for an agent to refactor and optimize React source code. Ingestion points: User-provided React source code. Boundary markers: Not specified. Capability inventory: Standard file system read/write via agent tools. Sanitization: N/A. While this involves processing untrusted code, the skill's instructions are defensive, focused solely on performance improvements, and do not introduce unsafe behaviors.
[DYNAMIC_EXECUTION]: One rule includes a code example using dangerouslySetInnerHTML to inject a small, synchronous theme-initialization script during server-side rendering. This is a standard and legitimate web development pattern used to prevent visual flickering (hydration mismatch) and does not constitute a security risk within the skill context.

react-best-practices