Database

Store your app's data in MongoDB collections that follow a schema you design, and get a ready API for every collection.

The Database module stores your app's data — orders, products, bookings — in MongoDB collections. You describe each collection once with a schema; Norbix then checks every write against it, builds the record form in Cloud, and serves a create, read, update, delete and query API for the collection. The data lives in MongoDB: your own cluster, or a managed Atlas Flex cluster that Norbix runs for you.

Enable the module

Open a project and choose Database in the left menu. While the module is off, Cloud shows this page; press ENABLE. You need a project role that can manage modules (for example Administrator).

Database → enable page, shown while the module is off.

After ENABLE, Cloud opens Database → Collections. A new project already has one database integration, Primary Database, so you can create a collection at once. To turn the module off again, see Settings — no data is deleted.

After ENABLE: the Collections screen and the Database menu.

Module map

Screen	What it is for
Collections	Your collections: their schemas, records, triggers and list columns and index warnings.
Schema	The schema contract for hand-written and AI-written schemas: the field types with every keyword and refusal, references, nested forms and arrays.
Taxonomies	Lists of terms (categories, tags, places) that records can point at, in a tree when needed.
Aggregates	Saved MongoDB aggregation pipelines that your app runs by id.
Integrations	The MongoDB databases your project stores its collections in.
Settings	Turn the module off for the project.

Your first collection

1. Check the integration — a new project has Primary Database as the default; connect your own MongoDB if you want the data in your cluster.

2. Create a schema — for example authors with a name, a country and a year of birth; press Save & publish.

3. Add a record — open the collection and press Add New.

4. Read it from code — find with the SDK, or norbix api database find with the CLI.

Use it from code

Every Do it from code block in the Database pages uses this client. Install @norbix.ai/ts (see SDK install) and set NORBIX_API_KEY to a project API key.

import { Norbix } from '@norbix.ai/ts';

const projectId = 'pr_4pTPFS01KwV4asdGqlVrIt';
const norbix = new Norbix({ projectId }); // the key comes from NORBIX_API_KEY

const { list } = await norbix.hub.database.getDatabaseSchemas();
console.log(list.items.map((s) => `${s.schemaName} v${s.latestVersion ?? '-'}`));
// [ 'books v1', 'reviews v-', 'authors v2' ]

The CLI lines use the generic norbix hub … and norbix api … commands of @norbix.ai/cli (signed in with norbix login, project id in NORBIX_PROJECT_ID):

norbix hub database schemas get

The client sends the project id with every call (the norbix-project-id header), so the requests do not repeat it. The CLI does the same with the project of norbix login --project …, norbix config set projectId … or NORBIX_PROJECT_ID. The short commands work too, for example norbix db find books and norbix db count books.

• The Hub (norbix.hub.database, norbix hub database …) manages the module: schemas, taxonomies, triggers, aggregates and integrations. It can also read and write records.

• The API (norbix.api.database, norbix api database …) is what your app calls: records, terms and saved aggregates. It also accepts a signed-in end user's token.

• Every record call can name a databaseIntegrationId; without it the default integration of the request's environment is used.

Related

Integrations concept · Triggers concept · SDK · CLI · API · What agents can do