Zod v4 Expert Guidance

Overview

This skill provides comprehensive guidance on using Zod v4, TypeScript's leading validation library. It covers breaking changes from v3, migration patterns, core API usage, and real-world validation patterns.

When to Use This Skill

Use this skill when:

• Writing validation schemas with Zod v4

• Migrating code from Zod v3 to v4

• Implementing form validation, API response validation, or config validation

• Working with TypeScript type inference from Zod schemas

• Debugging Zod validation errors

• Designing type-safe data models

• Creating recursive or complex nested schemas

Quick Reference

Task Reference Document

Migrating from Zod v3 references/migration-from-v3.md

Core API and primitives references/core-api.md

Real-world patterns references/common-patterns.md

Key Breaking Changes in v4

CRITICAL: Always review these when working with Zod v4:

• Error customization: Use error parameter instead of message , invalid_type_error , required_error

• String formats: Moved to top-level functions (e.g., z.email() not z.string().email() )

• Function schemas: Completely redesigned API using z.function({ input, output }).implement()

• Error handling: Use .issues instead of .errors , z.treeifyError() for formatting

• Object methods: Use z.strictObject() and z.looseObject() instead of .strict() and .passthrough()

• Extending schemas: Use shape spreading (z.object({ ...Base.shape, ... }) ) instead of .merge() or .extend()

• Records: Must specify both key and value schemas: z.record(keySchema, valueSchema)

• Defaults in optional fields: Now apply even inside optional fields (behavioral change)

• Number validation: Infinity rejected by default, .int() enforces safe range

See references/migration-from-v3.md for complete migration guide.

Common Patterns

Basic Validation

import { z } from 'zod/v4';

// Simple schema const userSchema = z.object({ name: z.string().min(1), email: z.email(), age: z.number().int().positive(), });

// Parse with error handling const result = userSchema.safeParse(data); if (!result.success) { console.error(result.error.issues); }

Type Inference

// Automatically infer TypeScript type from schema type User = z.infer<typeof userSchema>; // { name: string; email: string; age: number }

Custom Validation

const passwordSchema = z .string() .min(8, { error: 'Password must be at least 8 characters' }) .regex(/[A-Z]/, { error: 'Must contain uppercase letter' }) .regex(/[0-9]/, { error: 'Must contain number' });

Best Practices

• Always use .safeParse() for user input - Never let validation errors crash your app

• Leverage type inference - Don't manually type what Zod can infer

• Use top-level format validators - z.email() not z.string().email() (v4 pattern)

• Prefer z.strictObject()

• Catch typos and unexpected fields

• Keep refinements simple - Complex business logic should be separate

• Reuse schemas - Define once, reference everywhere

• Document complex schemas - Use TypeScript JSDoc comments

Workflow

When working with Zod:

• Define schemas - Start with the shape of your data

• Add validations - Layer on constraints (min, max, regex, etc.)

• Add custom errors - Make validation messages user-friendly

• Infer types - Use z.infer<typeof schema> for TypeScript types

• Parse safely - Use .safeParse() and handle errors gracefully

• Test edge cases - Validate your validation logic

Resources

• Migration Guide: references/migration-from-v3.md

• Core API Reference: references/core-api.md

• Common Patterns: references/common-patterns.md

• Official Docs: https://zod.dev/v4

Custom Instructions