A Runtime for Backend APIs

A backend runtime, not a codegen tool. Point FastBackend at a SQLAlchemy or Prisma schema — it serves CRUD routes, relationships, and OpenAPI at runtime. Routes live in memory, not on disk.

$ fastbackend init my-api --schema sqlalchemy
# → Scaffolds schema, config, and custom endpoints

$ fastbackend generate && fastbackend dev
# → Your API is live at http://localhost:8301/docs

Everything you need from a schema-driven backend

IR + OpenAPI on disk

The CLI writes `.fastbackend/ir.json` and OpenAPI from your schema. No CRUD route code is written to disk.

Runtime routes

FastAPI and Express adapters register CRUD, relationships, and validation at startup from IR.

Schema formats

SQLAlchemy and Prisma parsers produce the same IR. More formats planned.

Override system

Replace any runtime CRUD route with custom handlers using the `@override` decorator.

Frontend integration

OpenAPI output works with Orval, openapi-typescript, Hey API, or UIGen. Any frontend stack can consume the spec.

Docker ready

Scaffold Dockerfile templates with `fastbackend init --docker` and build with `docker:build`.

What's shipping

Core IR engine, FastAPI runtime, and CLI are ready for local development.

  • Core IR engine
  • SQLAlchemy parser
  • Prisma parser
  • FastAPI runtime
  • CLI (init, generate, dev)
  • Custom endpoints + overrides
  • Docs site
  • Express runtime (Prisma)
  • npm / PyPI publish
  • ·Declarative services (storage, OAuth)
  • ·Spring adapter