The Schema and Type System

What you'll walk away with

Explain the core ideas behind The Schema and Type System
Run the sample GraphQL query or code and verify its output
Apply the technique correctly to the Tutorial Platform and production scenarios

Build the mental model

A GraphQL schema is a contract that declares every shape of data the API can return, using types. Built-in scalar types are `String`, `Int`, `Float`, `Boolean`, and `ID`; `!` marks a field non-null, and syntax like `[Type!]!` means a non-null list of non-null items. `Query` is a special root type that declares every readable entry point. This type language is called Schema Definition Language (SDL).

Connect it to a real scenario

We declare the Tutorial Platform's core `Tutorial` type with `id`, `title`, `summary`, `lessonCount`, and `isFeatured` fields. Marking `id` as `ID!` (non-null) enforces at the schema level that every tutorial must have an id. Notice the difference inside `Query` between `tutorial(id: ID!): Tutorial` (a single, nullable item) and `tutorials: [Tutorial!]!` (a non-null list).

The schema is the API contract, with five built-in scalar types
`!` means non-null; `[Type!]!` is a non-null list of non-null items
`Query` is the root type declaring every readable entry point

Try the working example

graphql

type Tutorial {
  id: ID!
  title: String!
  summary: String
  lessonCount: Int!
  isFeatured: Boolean!
}

type Query {
  tutorial(id: ID!): Tutorial
  tutorials: [Tutorial!]!
}

You should see

You can read and write a simple schema using types, scalars, and non-null/list syntax.

5-minute try-it

Write an `Author` type with `id`, `name`, and a nullable `bio`, plus a `tutorials` list field, and a `Query` extension with `author(id: ID!): Author`.

One important caution

Marking a field non-null (`!`) and then having the resolver return `null` causes a GraphQL execution error. If a field can legitimately be missing, leave it nullable.

GraphQL — Schemas and Types — GraphQL

Easy traps

Marking a field non-null (`!`) and then having the resolver return `null` causes a GraphQL execution error. If a field can legitimately be missing, leave it nullable.
Validate sample queries and mutations on a local or test server with recoverable data before applying them to production.

Exercise

Write an `Author` type with `id`, `name`, and a nullable `bio`, plus a `tutorials` list field, and a `Query` extension with `author(id: ID!): Author`.

You'll know it worked when: You can read and write a simple schema using types, scalars, and non-null/list syntax.