Quick Start

Minimal config

Point GraphJin at your database in config/dev.yml. New source-aware projects should use sources::

sources:
  - name: app
    kind: database
    type: postgres
    default: true
    connection_string: ${DATABASE_URL}
    schema: public

Then start the service:

graphjin serve

Open the Web UI

Open http://localhost:8080/ after the service starts. The Runtime view is a quick visual check that GraphJin discovered the configured sources and is serving the built-in system roots.

Use the Workbench at http://localhost:8080/workbench to try queries interactively. It sends requests to the same /api/v1/graphql endpoint that applications and command-line smoke tests use.

First query

GraphJin discovers tables and relationships, then compiles GraphQL into database work.

query {
  products(limit: 3, order_by: { id: asc }) {
    id
    name
    owner {
      id
      fullName: full_name
    }
  }
}

The response shape follows the GraphQL selection. There are no user-written resolvers for database fields.

Call the endpoint

Keep a raw endpoint call in your smoke test even if you used the Workbench. This verifies the HTTP API path your application or integration will call.

curl http://localhost:8080/api/v1/graphql \
  -H 'content-type: application/json' \
  -d '{"query":"query { products(limit: 3, order_by: { id: asc }) { id name } }"}'

Pass variables as JSON. Keep filters inside the GraphQL query shape and pass leaf values through variables:

query Products($maxPrice: Float!) {
  products(where: { price: { lteq: $maxPrice } }, order_by: { price: asc }) {
    id
    name
    price
  }
}

{ "maxPrice": 25 }

Next steps

• Learn the query language.
• Add filters and cursor pagination.
• Lock down production with saved queries and RBAC.