config-schema-setup_skill

Overview

This skill provides a TypeScript-first, type-safe environment variable validation system using Zod with a Drizzle-like schema API. It organizes configuration into server-only and public fields, supports feature flags, either-or constraints, and includes client-side protection patterns. The goal is predictable runtime validation and clear compile-time types for full-stack AI apps. It ships as a set of production-ready patterns and step-by-step recipes.

How this skill works

You define a declarative schema using a Drizzle-style API that maps directly to Zod validators. The system generates strict runtime checks and infers TypeScript types so you get end-to-end type safety for env vars. It enforces server/public separation, validates mutually exclusive or either-or constraints, and produces client-safe subsets to avoid leaking secrets. Validation runs at startup and fails fast with descriptive errors.

When to use it

• When you need strict, typed validation of environment variables in a TypeScript full-stack app.
• When you must prevent accidental exposure of server-only secrets to the client.
• When feature flags or mutually exclusive config options require formal validation.
• When you want early-failure validation at app startup with readable error messages.
• When you want a consistent schema-driven approach across multiple services or deployments.

Best practices

• Declare separate server and public fields explicitly to avoid leaking secrets.
• Use Zod refinements for complex constraints and either-or rules for mutual exclusion.
• Keep feature flags as explicit booleans or typed enums and document their runtime behavior.
• Run validation during the application bootstrap to fail fast and provide actionable logs.
• Generate and export inferred TypeScript types from the schema for use throughout the codebase.

Example use cases

• A full-stack AI app that validates API keys and exposes only non-secret settings to the browser.
• A deployment pipeline that checks feature flags and either-or DB credentials before promoting releases.
• A monorepo where multiple services consume a shared schema with consistent type guarantees.
• A server-rendered app that must ensure environment shape consistency across staging and production.