Why use branded types instead of plain strings?

Branded types prevent logic errors where two different concepts (like a UserId and an OrderId) are both strings but should not be interchangeable in function arguments. It forces the developer to validate the string before using it as a specific type.

What is the 'Parse, Don't Validate' principle?

It is an architectural approach where you transform untrusted input into a richer, typed data structure that is guaranteed to be valid, rather than simply checking if data is valid and returning a boolean.

Where should validation happen according to this skill?

Validation should happen exclusively at the 'boundary'—the point where data enters your system from an untrusted source, such as an HTTP request, a CLI argument, or a database query result.

Does this skill support Express middleware?

Yes, it provides patterns for creating reusable Express middleware that validates request bodies against Zod schemas before passing control to the route handler.

Zod Validation Boundary

Name: Zod Validation Boundary
Author: jagreehal

byjagreehal

0•

API開発

Enforces secure data validation at system boundaries using Zod schemas and branded types to ensure type-safe business logic.

This skill implements the 'Parse, Don't Validate' architectural pattern to harden applications against malformed data. By establishing a strict validation layer at system boundaries—such as HTTP endpoints, CLI inputs, or message queues—it ensures that external data is transformed into rich, trusted types before reaching core business logic. This approach eliminates redundant checks within internal functions, prevents 'primitive obsession' through the use of branded types, and provides a standardized way to handle schema and domain-level validation errors in TypeScript environments.

主な機能

01Schema-based input validation using Zod for robust shape and format checking

02Implementation of Branded Types to prevent accidental mixing of similar primitive types like IDs

03Support for advanced patterns including type coercion, transforms, and partial updates

04Standardized validation error formatting for consistent API error responses

050 GitHub stars

06Clear separation between boundary schema validation and internal domain logic

ユースケース

01Hardening Express or Fastify API endpoints against invalid request bodies and query parameters

02Creating type-safe domain models where IDs and sensitive strings cannot be used interchangeably

03Refactoring legacy codebases to move validation logic from business services to the entry layer

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jagreehal/jagreehal-claude-skills validation-boundary

For use in Claude.ai and ChatGPT

Download Skill