engine

@launchstack/engine

The one-install aggregate: createEngine() assembles the feature packages behind a single CoreConfig, and re-exports their surfaces for consumers who want one dependency instead of eight. It deliberately does not contain logic of its own — everything here is composition and re-export.

Install

pnpm add @launchstack/engine

Use

import { createEngine } from "@launchstack/engine";

const engine = createEngine({
    db: { url: process.env.DATABASE_URL! },
    llm: {},
    embeddings: { indexName: "legacy-openai-1536" },
    ocr: { defaultProvider: "NATIVE_PDF" },
    providers: {},
    storage: myStoragePort,
});
await engine.close();

API

Subpath	What it is
.	createEngine + the aggregate re-export surface
./config	CoreConfig and its slices

Configuration

Nothing here reads process.env. Configuration is injected by the composition root — createEngine(config) in @launchstack/engine, or the package's own configure* hooks when used standalone.

The engine reads zero environment variables — all configuration flows through CoreConfig from the composition root (apps/web/src/server/engine.ts is the in-repo reference).

Stability

0.x. The aggregate exists for one-install consumers; fine-grained installs of the feature packages are the primary surface.

License

Apache-2.0 — see LICENSE.