The Agent Skills Directory

[SAFE]: The skill consists of markdown-based guidelines and code examples for optimizing React and Next.js applications. The content is focused on performance engineering and does not contain any executable scripts or malicious logic.
[EXTERNAL_DOWNLOADS]: The documentation references several well-known and trusted third-party libraries, including swr, lru-cache, and better-all. These are standard tools in the ecosystem and are referenced neutrally as part of the best-practice recommendations.
[PROMPT_INJECTION]: No evidence of prompt injection or behavior-override instructions was found. The skill includes specific security advice, such as recommending authentication within Server Actions, which improves the overall security posture of the applications it helps build.
[INDIRECT_PROMPT_INJECTION]: The skill establishes an attack surface as it is designed to process and refactor user-provided source code.
Ingestion points: The agent reads React and Next.js source code from the user.
Boundary markers: There are no explicit instructions or delimiters defined in the guidelines to separate user-provided code from the agent's internal instructions.
Capability inventory: The skill allows the agent to generate and refactor code. It also mentions environment setup commands like pnpm install in its documentation.
Sanitization: The skill does not provide specific sanitization or validation routines for the input code it processes.

vercel-react-best-practices