backend-api

You can expose a data model through GraphQL CRUD by using declarative @DataModel metadata, which automatically generates the create, read, update, and delete operations for your entities. This avoids manually writing resolvers for each endpoint.

Optimistic locking in a GraphQL backend API prevents concurrent update conflicts by checking version consistency before applying mutations. It returns specific optimistic locking conflict errors within a consistent API error response shape when data versions mismatch.

The best way to handle validation errors is by using validation-aware union handling and ExpectedError unions. This structure returns framework validation errors and expected business errors in a consistent, strongly typed API response format.

You can add typed custom actions to a GraphQL backend by declaring @Action metadata with strongly typed parameters and return values. This allows you to implement domain-specific endpoints with clear success or failure unions.

Yes, you can enforce permission checks before executing a custom GraphQL action by applying canExecute guards. These guards validate user permissions and state checks before the action processes, returning user-friendly errors upon failure.

You need to use ExpectedError unions in GraphQL API design when you want to consistently distinguish between successful operations and anticipated business failures. This approach ensures validation flows and expected errors are strongly typed alongside successful return values.