The Agent Skills Directory

[SAFE]: The skill provides high-quality technical guidance focused on performance and security best practices. It includes specific rules for authenticating server-side mutations and preventing concurrency-related data leaks.
[EXTERNAL_DOWNLOADS]: The guidelines recommend established community packages such as swr, lru-cache, and better-all. These references are provided in the context of standard optimization patterns and do not involve suspicious download-and-execute behaviors.
[DATA_EXPOSURE_AND_EXFILTRATION]: No evidence of hardcoded credentials, sensitive file access, or unauthorized network operations was found. The code examples provided are for educational and refactoring purposes within a development environment.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze and refactor user-provided React/Next.js code, which is a potential surface for indirect prompt injection if the ingested data contains malicious instructions.
Ingestion points: React components and Next.js project files provided by the user (as defined in the SKILL.md triggers).
Boundary markers: The skill does not specify the use of boundary markers or instructions to ignore embedded content when processing user code.
Capability inventory: The skill guides automated refactoring and code generation, which typically involves file-system access and command execution in the agent's environment.
Sanitization: No specific sanitization or validation logic for the content of user-provided code is included in these guidelines.

vercel-react-best-practices