JSON Schema to TypeScript

Overview

You send a JSON Schema and get TypeScript source back, byte-identical every time for the same input. Nested objects become their own named types, enums and const become string-literal unions, anyOf and oneOf become unions, allOf becomes an intersection, and a local $ref into $defs or definitions points at the type we generated for it. Whatever TypeScript cannot express comes out as unknown, flagged by lossy and named in warnings — we never quietly invent a type. There is no runtime validator in the output, so pair it with schema.validate_json if you need checks at runtime.

Hand-written types drift away from the contract the moment someone edits the schema. You regenerate them in one call, with no build toolchain to install.

When to use it

Scope

Non-goals

• We never download a $ref from another document; keep them local
• We never compile, format, or run the TypeScript we hand you
• We never emit a runtime validator, only compile-time types
• We cover a subset, not a full converter: if and then and else, dependentSchemas, patternProperties, prefixItems nuances, and pattern or format constraints come out as unknown with a warning rather than a clever approximation

Properties

Capabilities

Cost & billing

Flat rate per invocation

File handling

How we handle your data

Input schema

Output schema

API contract

curl -sS -X POST "$BASE/v1/execute" -H "Authorization: Bearer $KEY" -H 'Content-Type: application/json' -d '{"skillKey":"schema.convert_json_schema_to_typescript","input":{"schema":{"type":"object","properties":{"sku":{"type":"string"},"quantity":{"type":"integer"}},"required":["sku"]},"rootTypeName":"OrderLine","style":"interface","additionalPropertiesMode":"indexSignature"}}'

import { Squillify } from "@squillify/sdk";

const sq = new Squillify({ apiKey: process.env.SQUILLIFY_API_KEY });

const { output } = await sq.execute("schema.convert_json_schema_to_typescript", {"rootTypeName":"OrderLine","schema":{"type":"object","properties":{"sku":{"type":"string","description":"Stock keeping unit"},"quantity":{"type":"integer"}},"required":["sku"]}});

Errors

Other limits

Limitations

• We never download a $ref from another document; keep them local
• We never compile, format, or run the TypeScript we hand you
• We never emit a runtime validator, only compile-time types
• We cover a subset, not a full converter: if and then and else, dependentSchemas, patternProperties, prefixItems nuances, and pattern or format constraints come out as unknown with a warning rather than a clever approximation
• Schemas that lean on remote references or conditional subschemas
• Getting a runtime validator rather than compile-time types
• Producing types for a language other than TypeScript

How to call

POST /v1/execute (requires authentication)

{
  "skillKey": "schema.convert_json_schema_to_typescript",
  "input": {
    "rootTypeName": "OrderLine",
    "schema": {
      "type": "object",
      "properties": {
        "sku": {
          "type": "string",
          "description": "Stock keeping unit"
        },
        "quantity": {
          "type": "integer"
        }
      },
      "required": [
        "sku"
      ]
    }
  }
}

Examples

{
  "rootTypeName": "OrderLine",
  "schema": {
    "type": "object",
    "properties": {
      "sku": {
        "type": "string",
        "description": "Stock keeping unit"
      },
      "quantity": {
        "type": "integer"
      }
    },
    "required": [
      "sku"
    ]
  }
}

{
  "warnings": [],
  "lossy": false,
  "emittedRootTypeName": "OrderLine",
  "declarations": [
    "OrderLine"
  ],
  "typescript": "export interface OrderLine {\n  /** Stock keeping unit */\n  sku: string;\n  quantity?: number;\n}\n"
}

{
  "schema": {
    "type": "string",
    "enum": [
      "draft",
      "issued",
      "paid"
    ]
  },
  "rootTypeName": "InvoiceStatus",
  "emitComments": false
}

{
  "warnings": [],
  "lossy": false,
  "emittedRootTypeName": "InvoiceStatus",
  "declarations": [
    "InvoiceStatus"
  ],
  "typescript": "export type InvoiceStatus = \"draft\" | \"issued\" | \"paid\";\n"
}

Recommended use cases

• Generate client-side types from the JSON Schema a partner publishes
• Scaffold TypeScript models for a config file you already describe with a schema
• Keep front-end types in step with a contract without adding a build toolchain

Avoid for

• Schemas that lean on remote references or conditional subschemas
• Getting a runtime validator rather than compile-time types
• Producing types for a language other than TypeScript

Related skills

Composable with

Safety & privacy

• Runs without calling any external service.
• Your payloads are never stored permanently.
• Telemetry captures metadata only — never your payload contents.
• We never train models on your payloads.
• No human reviews your payloads.

Tags