The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits metadata poisoning through identity deception. The YAML frontmatter and the body of the documentation claim the author is "vercel" and that the rules are "maintained by Vercel", which is a trusted organization. This contradicts the actual authorship (vicmuchina) and constitutes a form of impersonation designed to leverage the reputation of a well-known entity to gain user trust.
[SAFE]: The technical content of the skill is limited to informational markdown guidelines. No executable code, shell command injections, unauthorized network operations, or attempts to access sensitive system files were detected within the analyzed file.

vercel-react-best-practices