bb-graphql-pothos

To keep GraphQL resolvers thin, standardize endpoint construction by ensuring resolvers only map parameters and return results while delegating all business logic to a dedicated service layer.

The best way to structure GraphQL queries and mutations with Pothos and Yoga is to enforce a clear separation of concerns, keeping inputs and type definitions outside resolver logic and delegating execution to services.

Yes, you can use this architecture to add or update Queries, Mutations, and Type definitions across multiple APIs, specifically including bruno-brown-bk, internal-api, and orders-api.

GraphQL resolvers should avoid direct database access to prevent endpoints from becoming bloated, ensuring that services handle the core logic and resolvers remain focused on parameter mapping.

You need a service layer for GraphQL endpoint construction when resolvers begin mixing business logic with data access, requiring services to handle the core logic while resolvers strictly delegate execution.