JSON Schema Generator Tutorial with Examples

Writing a JSON Schema by hand is tedious and error-prone, especially for deeply nested API responses. A JSON Schema generator takes an example JSON payload and produces a schema automatically, which you can then use for validation, documentation, and testing. This tutorial walks through generating and refining a schema from a real example.

Starting with example JSON

Suppose you have an API endpoint that returns a user profile:

{
  "id": 1042,
  "name": "Priya Sharma",
  "email": "priya.sharma@example.com",
  "isActive": true,
  "roles": ["admin", "editor"],
  "address": {
    "city": "Bengaluru",
    "pincode": "560001"
  }
}

The generated JSON Schema

A JSON Schema generator inspects each field's type and structure and produces a Draft-07 (or newer) compliant schema like this:

{
  "$schema": "http://json-schema.org/draft-07/schema#",
  "type": "object",
  "properties": {
    "id": { "type": "integer" },
    "name": { "type": "string" },
    "email": { "type": "string" },
    "isActive": { "type": "boolean" },
    "roles": {
      "type": "array",
      "items": { "type": "string" }
    },
    "address": {
      "type": "object",
      "properties": {
        "city": { "type": "string" },
        "pincode": { "type": "string" }
      },
      "required": ["city", "pincode"]
    }
  },
  "required": ["id", "name", "email", "isActive", "roles", "address"]
}

Refining the generated schema

Auto-generated schemas are a starting point, not a final product. You typically need to add stricter validation rules manually:

• Add "format": "email" to the email field for format-level validation.
• Add "pattern" for the pincode field to enforce a 6-digit Indian PIN code, e.g. ^[0-9]{6}$.
• Move optional fields out of the "required" array if they are not always present.
• Add "enum" to the roles items if only specific role values are allowed.
• Set "additionalProperties": false to reject unexpected extra fields.

Here is the address sub-schema after refinement:

"address": {
  "type": "object",
  "properties": {
    "city": { "type": "string", "minLength": 1 },
    "pincode": { "type": "string", "pattern": "^[0-9]{6}$" }
  },
  "required": ["city", "pincode"],
  "additionalProperties": false
}

Using the schema for validation

Once generated, the schema can be used with a validation library like Ajv in Node.js to check incoming data before it reaches your business logic:

const Ajv = require('ajv');
const ajv = new Ajv();
const validate = ajv.compile(userSchema);

const valid = validate(incomingUserData);
if (!valid) {
  console.log(validate.errors);
}

This same schema can also be reused to generate API documentation (via OpenAPI/Swagger) and to auto-generate TypeScript interfaces, avoiding duplicate work across your codebase.

Frequently Asked Questions