vercel-react-v1
Pass
Audited by Gen Agent Trust Hub on Aug 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONMETADATA_POISONINGDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: As a coding assistant skill, this tool ingests and refactors user-provided React and Next.js source code, which serves as a potential vector for untrusted data injection. * Ingestion points: React and Next.js source code files provided by the user in the agent's context. * Boundary markers: The skill does not define specific delimiters for user code, relying on the host agent's default safety mechanisms. * Capability inventory: The skill instructs the agent to generate code that performs network fetches, file system reads (fs.readFileSync), and dynamic script injection. * Sanitization: The guidelines explicitly recommend using Zod for input validation and implementing mandatory authentication/authorization checks in server-side logic.
- [METADATA_POISONING]: There is a discrepancy between the author field in the skill manifest ('vercel') and the attribution provided in the system metadata ('blockmatic'). This is documented as a metadata inconsistency; however, the content accurately reflects Vercel's public technical documentation.
- [DYNAMIC_EXECUTION]: The skill recommends the use of dangerouslySetInnerHTML to inject synchronous scripts for preventing hydration flickering. While this is a sensitive React capability, it is documented here as a legitimate performance optimization pattern with clear, safe implementation examples.
Audit Metadata