vercel-react-best-practices

Pass

Audited by Gen Agent Trust Hub on Mar 8, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious behavior or security threats were identified. The skill content is purely instructional documentation focused on web performance optimization.- [SAFE]: The skill metadata lists 'vercel' as the author while the actual provider is 'fimoklei'. This is interpreted as a legitimate source attribution for the documented best practices rather than a deceptive impersonation attempt.- [PROMPT_INJECTION]: The skill serves as a comprehensive instruction set designed to influence the agent's code generation behavior. This constitutes a surface for indirect prompt injection.
  • Ingestion points: Technical rules and best practices are ingested from AGENTS.md and the markdown files within the rules/ directory.
  • Boundary markers: Absent. The skill does not provide specific delimiters to isolate its instructions from other contextual data.
  • Capability inventory: The instructions guide the agent in performing code refactoring and generation, which involve file system operations (write) and potential build process execution.
  • Sanitization: While not directly performing data processing, the guidelines themselves promote security best practices such as input validation via Zod and manual authorization checks in server actions.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 8, 2026, 06:47 PM