The Agent Skills Directory

[SAFE]: The skill is a collection of 65 performance optimization rules for React and Next.js, including detailed explanations and good/bad code examples. No malicious instructions or patterns were detected in the documentation or rule files.- [EXTERNAL_DOWNLOADS]: The documentation references and recommends several well-known and trusted external libraries such as swr, better-all, and lru-cache. It also refers to official documentation from React, Next.js, and Vercel. These references are used for educational purposes to illustrate best practices.- [INDIRECT_PROMPT_INJECTION]: The skill provides a set of patterns that an agent would use to analyze and refactor user-provided React/Next.js code. This constitutes an attack surface for indirect prompt injection if the user-supplied code contains hidden instructions designed to manipulate the agent.
Ingestion points: The agent ingests user-supplied source code from files (e.g., .tsx, .ts, .js).
Boundary markers: The skill does not provide specific delimiters or 'ignore embedded instructions' markers to wrap user code during analysis.
Capability inventory: The skill itself contains no active code execution capabilities (e.g., subprocess, eval, or network operations).
Sanitization: No sanitization or validation of the input source code is performed by the skill guidelines.- [NO_CODE]: The skill consists entirely of markdown documentation files. It does not include any executable scripts (.js, .py, .sh), binaries, or configuration files that trigger command execution in the agent's environment.

vercel-react-best-practices