Skip to main content

Code Generation

Transform your (synced & normalized) OpenAPI/Swagger specs into a ready‑to‑use SDK composed of framework‑agnostic data types and framework‑specific integration logic.

What “Generate” Does​

  • Reads normalized specs from .intrig/specs/{sourceId}-latest.json.

  • Produces two categories of artifacts per source:

    • Data Types – framework‑agnostic.
    • Integration Logic – framework‑specific.

  • Publishes the SDK into your project’s node_modules for immediate import.

Always run Synchronization first so specs are current.

Command​

intrig generate

Generates for all configured sources. If no normalized file exists, run intrig sync first.

Outputs​

  1. Data Types (framework‑agnostic)

    • TypeScript types/interfaces – DTOs, enums, schemas.
    • Zod schemas – runtime validation aligned with types.
    • JSON Schema – machine‑readable format for tooling.

  2. Integration Logic (framework‑specific)

    • Generated per selected framework.
    • Example: React hooks with [state, call] signature.
    • Directory structure and import paths are framework‑specific; see the corresponding framework documentation for details.

When to Run It​

  • At a fresh clone of the repository.
  • After a sync operation.
  • After switching branches.
  • After running npm install (since it can wipe out the generated SDK).
  • When backend APIs change (endpoints, params, schemas).
  • In CI, so the SDK is available for the application to compile.

Regeneration & Overwrites​

  • Overwrites published SDK in node_modules.
  • Do not edit generated files; changes will be lost.

Gochas (Common Pitfalls)​

  • Imports break → operationId changed. Search by Tag or path to update.
  • Type errors → Schema changed. Update call sites.
  • Spec missing → Run intrig sync first.

See Also​

  • Synchronization – fetching & normalizing specs.
  • Framework Guides – integration logic per framework.