The Agent Skills Directory

[SAFE] (SAFE): Extensive analysis of all 58 files (primarily markdown rules and TypeScript build scripts) confirms the skill is safe. No indicators of prompt injection, data exfiltration, or obfuscation were found.- [COMMAND_EXECUTION] (SAFE): The skill includes local build and validation scripts in the src/ directory (e.g., build.ts, validate.ts). These scripts use the fs/promises module to read markdown files from the rules/ directory and compile them into a single guide. No arbitrary or dangerous command execution was detected.- [EXTERNAL_DOWNLOADS] (SAFE): References to external libraries (e.g., better-all, lru-cache, swr) are for documentation purposes and do not involve unauthorized background downloads during skill execution. The package.json only includes standard development dependencies used for local processing.

vercel-react-best-practices