The Agent Skills Directory

[SAFE]: The skill serves as a detailed security guideline for auditing and fixing vulnerabilities in React, Express, and Next.js applications.
[COMMAND_EXECUTION]: Includes instructions for using standard shell commands such as git ls-files to check for tracked environment files and npm run build for project verification.
[EXTERNAL_DOWNLOADS]: References Vercel's official API endpoints for managing environment variables. Interacting with this well-known service for project configuration is considered safe practice in this context.
[DATA_EXFILTRATION]: No unauthorized data exfiltration was detected. The network interactions are scoped to authenticated configuration of the user's Vercel project.
[PROMPT_INJECTION]: The skill does not contain instructions to bypass safety filters or override agent behavior. The parallel audit agent prompts are strictly focused on security analysis tasks.

vibe-proof