vercel-react-best-practices
Pass
Audited by Gen Agent Trust Hub on Apr 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves as a technical documentation and rule-set for optimizing React and Next.js code. It does not perform any dangerous operations, request sensitive permissions, or contain malicious instructions.
- [EXTERNAL_DOWNLOADS]: The skill references several external resources for documentation and utility purposes, including official React (react.dev), Next.js (nextjs.org), and SWR (swr.vercel.app) websites, as well as Vercel's blog and public GitHub repositories (e.g., shuding/better-all). These are all well-known and trusted sources associated with the skill's primary purpose.
- [REMOTE_CODE_EXECUTION]: While the skill's instructions guide an agent to refactor code, it does not itself execute remote scripts or download untrusted binaries. The build and validation scripts mentioned in the README are standard development utilities for managing the rule-set.
- [DATA_EXFILTRATION]: No patterns of data exfiltration were found. The guidelines actually include security-positive rules, such as
server-auth-actions.mdwhich emphasizes authenticating Server Actions, andserver-no-shared-module-state.mdwhich warns against data leaks in concurrent environments. - [PROMPT_INJECTION]: The skill contains no instructions designed to bypass safety filters or override the agent's core behavioral constraints. The directive to follow the rules for automation is consistent with the skill's function as a developer tool.
- [INDIRECT_PROMPT_INJECTION]: As a code-analysis and refactoring skill, it inherently processes user-provided React/Next.js code.
- Ingestion points: Reads user-supplied source code files.
- Boundary markers: Absent in the static rules themselves.
- Capability inventory: Primarily involves code analysis and transformation; typical agent platforms restrict file system and network access for such tasks.
- Sanitization: Not applicable to static rule content.
- This represents a standard surface for code-processing tasks and is not a finding against the skill's integrity.
Audit Metadata