type-safety-validation_skill

Overview

This skill teaches end-to-end type safety using Zod for runtime validation, tRPC for type-safe APIs, Prisma as a typed ORM, and modern TypeScript features (5.7+). It helps you catch errors at compile time, validate external input at boundaries, and keep consistent types from the database through the API to the UI. Ideal for full‑stack apps and migrations from JavaScript to strict TypeScript.

How this skill works

You define Zod schemas to validate and transform runtime data, infer TypeScript types from those schemas, and reuse them across layers. tRPC exposes procedures whose inputs and outputs are validated by Zod and remain fully typed on the client. Prisma generates typed database models and query helpers that integrate with the validated shapes. TypeScript 5.7+ language features are used to keep literal types, guarantees, and ergonomics consistent across the stack.

When to use it

• Building full‑stack applications that require strict typing end‑to‑end
• Designing APIs where client and server must share exact types (tRPC, RPC, GraphQL)
• Validating user input, webhooks, or any external data at service boundaries
• Ensuring database queries and mutations are statically typed and safe
• Migrating a JavaScript codebase to TypeScript with runtime validation
• Implementing strict security or compliance rules around input validation

Best practices

• Validate at boundaries: API inputs, form submissions, and external payloads
• Prefer .safeParse() for graceful error handling; use .parse() when data is already trusted
• Reuse Zod schemas to avoid duplicate runtime/runtime types — infer types with z.infer
• Enable strict TypeScript settings (strict, noUncheckedIndexedAccess) and prefer unknown over any
• Brand IDs and use discriminated unions for variant shapes; cache or memoize heavy schemas
• Batch tRPC calls and optimize Prisma queries; validate env vars at startup

Example use cases

• A Next.js app using tRPC + Prisma where client code has exact types for queries and mutations
• A webhook handler that validates incoming payloads with Zod before writing to the database
• A migration plan to convert a JS codebase to TS by introducing Zod schemas and incremental typing
• A multi‑service system where DTOs are validated and shared between services via tRPC
• Form validation in React where Zod schemas drive both UI validation and server checks