Amazon Ion Schema defines a grammar and constraints for narrowing the universe of Ion values. A schema consists of zero or more types, and a type is a collection of zero or more constraints over the Ion data model. Aspects of a value not constrained by a type (“open content”) are considered valid, which enables loosely-coupled systems to evolve independently.

Once defined, a type can be used to:

For more information, see the Ion Schema Specification.

Latest News

🚀 Launch Announcement 🚀 Transitive Import Repair Tool
07 June 2023

The CLI in ion-schema-kotlin now includes a command for repairing schemas that rely on the bug that allowed transitive imports.

Read more

Ion Schema Rust 0.5.0 Released
13 September 2022

Ion Schema Rust 0.5.0 is now available. This beta release includes all the functionalities of Ion Schema as per the Ion Schema 1.0 specification.

Read more

New tool: Ion Schema Sandbox
12 September 2022

A browser-based sandbox environment for Ion Schema has been added to the Ion Schema website.

Read more

Visit the News page for more announcements related to Ion Schema.

Ion Schema Example

  name: Person,
  type: struct,
  fields: {
    title: {
      type: symbol,
      valid_values: [Mr, Mrs, Miss, Ms, Mx, Dr],
    firstName: { type: string, occurs: required },
    middleName: string,
    lastName: { type: string, occurs: required },
    age: { type: int, valid_values: range::[0, 130] },

The following values are valid for the type Person:

  firstName: "Susan",
  lastName: "Jones",
  title: Mr,
  firstName: "Jonah",
  middleName: "Q.",
  lastName: "Smith",
  age: 34,

The following values are not valid for the type Person:

  firstName: "Cathy",        // lastName is required
  title: Prof,               // Prof is not a valid value for the title field
  firstName: "Jasmine",
  lastName: "Bradford",
  firstName: "Shami",
  lastName: "Ahmed",
  age: 131,                  // age must be between 0 and 130, inclusive