AGENT SKILL

zod-validation-expert

Expert in Zod — TypeScript-first schema validation. Covers parsing, custom errors, refinements, type inference, and integration with React Hook Form, Next.js, and tRPC.

License 许可
MIT
最近更新
2026-05-10
Stars
★ 1
Forks
1

01 前言 Front Matter

name
zod-validation-expert
risk
safe
source
community
date_added
2026-03-05
description
Expert in Zod — TypeScript-first schema validation. Covers parsing, custom errors, refinements, type inference, and integration with React Hook Form, Next.js, and tRPC.

03 SKILL.md 正文

Zod Validation Expert

You are a production-grade Zod expert. You help developers build type-safe schema definitions and validation logic. You master Zod fundamentals (primitives, objects, arrays, records), type inference (z.infer), complex validations (.refine, .superRefine), transformations (.transform), and integrations across the modern TypeScript ecosystem (React Hook Form, Next.js API Routes / App Router Actions, tRPC, and environment variables).

When to Use This Skill

  • Use when defining TypeScript validation schemas for API inputs or forms
  • Use when setting up environment variable validation (process.env)
  • Use when integrating Zod with React Hook Form (@hookform/resolvers/zod)
  • Use when extracting or inferring TypeScript types from runtime validation schemas
  • Use when writing complex validation rules (e.g., cross-field validation, async validation)
  • Use when transforming input data (e.g., string to Date, string to number coercion)
  • Use when standardizing error message formatting

Core Concepts

Why Zod?

Zod eliminates the duplication of writing a TypeScript interface and a runtime validation schema. You define the schema once, and Zod infers the static TypeScript type. Note that Zod is for parsing, not just validation. safeParse and parse return clean, typed data, stripping out unknown keys by default.

Schema Definition & Inference

Primitives & Coercion

import { z } from "zod";

// Basic primitives
const stringSchema = z.string().min(3).max(255);
const numberSchema = z.number().int().positive();
const dateSchema = z.date();

// Coercion (automatically casting inputs before validation)
// Highly useful for FormData in Next.js Server Actions or URL queries
const ageSchema = z.coerce.number().min(18); // "18" -> 18
const activeSchema = z.coerce.boolean(); // "true" -> true
const dobSchema = z.coerce.date(); // "2020-01-01" -> Date object

Objects & Type Inference

const UserSchema = z.object({
  id: z.string().uuid(),
  username: z.string().min(3).max(20),
  email: z.string().email(),
  role: z.enum(["ADMIN", "USER", "GUEST"]).default("USER"),
  age: z.number().min(18).optional(), // Can be omitted
  website: z.string().url().nullable(), // Can be null
  tags: z.array(z.string()).min(1), // Array with at least 1 item
});

// Infer the TypeScript type directly from the schema
// No need to write a separate `interface User { ... }`
export type User = z.infer<typeof UserSchema>;

Advanced Types

// Records (Objects with dynamic keys but specific value types)
const envSchema = z.record(z.string(), z.string()); // Record<string, string>

// Unions (OR)
const idSchema = z.union([z.string(), z.number()]); // string | number
// Or simpler:
const idSchema2 = z.string().or(z.number());

// Discriminated Unions (Type-safe switch cases)
const ActionSchema = z.discriminatedUnion("type", [
  z.object({ type: z.literal("create"), id: z.string() }),
  z.object({ type: z.literal("update"), id: z.string(), data: z.any() }),
  z.object({ type: z.literal("delete"), id: z.string() }),
]);

Parsing & Validation

parse vs safeParse

const schema = z.string().email();

// ❌ parse: Throws a ZodError if validation fails
try {
  const email = schema.parse("invalid-email");
} catch (err) {
  if (err instanceof z.ZodError) {
    console.error(err.issues);
  }
}

// ✅ safeParse: Returns a result object (No try/catch needed)
const result = schema.safeParse("user@example.com");

if (!result.success) {
  // TypeScript narrows result to SafeParseError
  console.log(result.error.format()); 
  // Early return or throw domain error
} else {
  // TypeScript narrows result to SafeParseSuccess
  const validEmail = result.data; // Type is `string`
}

Customizing Validation

Custom Error Messages

const passwordSchema = z.string()
  .min(8, { message: "Password must be at least 8 characters long" })
  .max(100, { message: "Password is too long" })
  .regex(/[A-Z]/, { message: "Password must contain at least one uppercase letter" })
  .regex(/[0-9]/, { message: "Password must contain at least one number" });

// Global custom error map (useful for i18n)
z.setErrorMap((issue, ctx) => {
  if (issue.code === z.ZodIssueCode.invalid_type) {
    if (issue.expected === "string") return { message: "This field must be text" };
  }
  return { message: ctx.defaultError };
});

Refinements (Custom Logic)

// Basic refinement
const passwordCheck = z.string().refine((val) => val !== "password123", {
  message: "Password is too weak",
});

// Cross-field validation (e.g., password matching)
const formSchema = z.object({
  password: z.string().min(8),
  confirmPassword: z.string()
}).refine((data) => data.password === data.confirmPassword, {
  message: "Passwords don't match",
  path: ["confirmPassword"], // Sets the error on the specific field
});

Transformations

// Change data during parsing
const stringToNumber = z.string()
  .transform((val) => parseInt(val, 10))
  .refine((val) => !isNaN(val), { message: "Not a valid integer" });

// Now the inferred type is `number`, not `string`!
type TransformedResult = z.infer<typeof stringToNumber>; // number

Integration Patterns

React Hook Form

import { useForm } from "react-hook-form";
import { zodResolver } from "@hookform/resolvers/zod";
import { z } from "zod";

const loginSchema = z.object({
  email: z.string().email("Invalid email address"),
  password: z.string().min(6, "Password must be 6+ characters"),
});

type LoginFormValues = z.infer<typeof loginSchema>;

export function LoginForm() {
  const { register, handleSubmit, formState: { errors } } = useForm<LoginFormValues>({
    resolver: zodResolver(loginSchema)
  });

  const onSubmit = (data: LoginFormValues) => {
    // data is fully typed and validated
    console.log(data.email, data.password);
  };

  return (
    <form onSubmit={handleSubmit(onSubmit)}>
      <input {...register("email")} />
      {errors.email && <span>{errors.email.message}</span>}
      {/* ... */}
    </form>
  );
}

Next.js Server Actions

"use server";
import { z } from "zod";

// Coercion is critical here because FormData values are always strings
const createPostSchema = z.object({
  title: z.string().min(3),
  content: z.string().optional(),
  published: z.coerce.boolean().default(false), // checkbox -> "on" -> true
});

export async function createPost(prevState: any, formData: FormData) {
  // Convert FormData to standard object using Object.fromEntries
  const rawData = Object.fromEntries(formData.entries());
  
  const validatedFields = createPostSchema.safeParse(rawData);
  
  if (!validatedFields.success) {
    return {
      errors: validatedFields.error.flatten().fieldErrors,
    };
  }
  
  // Proceed with validated database operation
  const { title, content, published } = validatedFields.data;
  // ...
  return { success: true };
}

Environment Variables

// Make environment variables strictly typed and fail-fast
import { z } from "zod";

const envSchema = z.object({
  DATABASE_URL: z.string().url(),
  NODE_ENV: z.enum(["development", "test", "production"]).default("development"),
  PORT: z.coerce.number().default(3000),
  API_KEY: z.string().min(10),
});

// Fails the build immediately if env vars are missing or invalid
const env = envSchema.parse(process.env);

export default env;

Best Practices

  • ✅ Do: Co-locate schemas alongside the components or API routes that use them to maintain separation of concerns.
  • ✅ Do: Use z.infer<typeof Schema> everywhere instead of maintaining duplicate TypeScript interfaces manually.
  • ✅ Do: Prefer safeParse over parse to avoid scattered try/catch blocks and leverage TypeScript's control flow narrowing for robust error handling.
  • ✅ Do: Use z.coerce when accepting data from URLSearchParams or FormData, and be aware that z.coerce.boolean() converts standard "false"/"off" strings unexpectedly without custom preprocessing.
  • ✅ Do: Use .flatten() or .format() on ZodError objects to easily extract serializable, human-readable errors for frontend consumption.
  • ❌ Don't: Rely exclusively on .partial() for update schemas if field types or constraints differ between creation and update operations; define distinct schemas instead.
  • ❌ Don't: Forget to pass the path option in .refine() or .superRefine() when performing object-level cross-field validations, otherwise the error won't attach to the correct input field.

Troubleshooting

Problem: Type instantiation is excessively deep and possibly infinite. Solution: This occurs with extreme schema recursion (e.g. deeply nested self-referential schemas). Use z.lazy(() => NodeSchema) for recursive structures and define the base TypeScript type explicitly instead of solely inferring it.

Problem: Empty strings pass validation when using .optional(). Solution: .optional() permits undefined, not empty strings. If an empty string means "no value," use .or(z.literal("")) or preprocess it: z.string().transform(v => v === "" ? undefined : v).optional().

Limitations

  • Use this skill only when the task clearly matches the scope described above.
  • Do not treat the output as a substitute for environment-specific validation, testing, or expert review.
  • Stop and ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.

04 同仓库更多 Skills

Sansar35
Sansar35/claude-agent

zustand-store-ts

Create Zustand stores following established patterns with proper TypeScript types and middleware.

★ 1
Sansar35
Sansar35/claude-agent

zod-validation-expert

Expert in Zod — TypeScript-first schema validation. Covers parsing, custom errors, refinements, type inference, and integration with React Hook Form, Next.js, and tRPC.

★ 1
Sansar35
Sansar35/claude-agent

zinc-database

Access ZINC (230M+ purchasable compounds). Search by ZINC ID/SMILES, similarity searches, 3D-ready structures for docking, analog discovery, for virtual screening and drug discovery.

★ 1
Sansar35
Sansar35/claude-agent

zarr-python

Chunked N-D arrays for cloud storage. Compressed arrays, parallel I/O, S3/GCS integration, NumPy/Dask/Xarray compatible, for large-scale scientific computing pipelines.

★ 1
Sansar35
Sansar35/claude-agent

zapier-workflows

Manage and trigger pre-built Zapier workflows and MCP tool orchestration. Use when user mentions workflows, Zaps, automations, daily digest, research, search, lead tracking, expenses, or asks to "run" any process. Also handles Perplexity-based research and Google Sheets data tracking.

★ 1
Sansar35
Sansar35/claude-agent

zapier-make-patterns

No-code automation democratizes workflow building. Zapier and Make (formerly Integromat) let non-developers automate business processes without writing code. But no-code doesn't mean no-complexity - these platforms have their own patterns, pitfalls, and breaking points. This skill covers when to use which platform, how to build reliable automations, and when to graduate to code-based solutions. Key insight: Zapier optimizes for simplicity and integrations (7000+ apps), Make optimizes for power

★ 1