The Agent Skills Directory

[METADATA_POISONING]: The skill's YAML frontmatter in SKILL.md identifies the author as vercel, which creates a deceptive mismatch with the actual skill author kimny1143.
[EXTERNAL_DOWNLOADS]: The documentation in AGENTS.md and rules/async-dependencies.md recommends utilizing the third-party library better-all from the shuding GitHub account.
[EXTERNAL_DOWNLOADS]: The documentation in rules/server-cache-lru.md recommends the implementation of the lru-cache library from the isaacs GitHub account.
[DYNAMIC_EXECUTION]: The rule described in rules/rendering-hydration-no-flicker.md provides implementation instructions for using dangerouslySetInnerHTML to execute a synchronous inline <script> that reads data from localStorage at the time of hydration.
[COMMAND_EXECUTION]: The rule in rules/rendering-svg-precision.md suggests the execution of npx svgo to perform coordinate precision reduction on SVG files.
[INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for processing and refactoring user-provided React and Next.js source code.
Ingestion points: React components and Next.js page files provided by the user for review or refactoring as described in SKILL.md.
Boundary markers: There are no explicit instructions or delimiters provided to ensure the agent ignores malicious instructions that might be embedded in code comments or string literals within the user's files.
Capability inventory: The skill enables the agent to perform automated refactoring, code generation, and bundle optimization tasks.
Sanitization: The instructions do not define any sanitization or validation steps for the input source code before it is incorporated into the agent's context for processing.

vercel-react-best-practices