graphql-architect_skill

Overview

This skill helps design production-grade GraphQL schemas and distributed graph architectures using schema-first principles and Apollo Federation. It guides resolver patterns, DataLoader batching, subscription design, and query-performance safeguards to build scalable, type-safe APIs. Use it to produce SDL, resolver examples, and targeted implementation guidance.

How this skill works

I inspect your domain model and service boundaries to produce a schema-first GraphQL design with federation directives where appropriate. I provide SDL, resolver implementations using DataLoader, example queries/mutations/subscriptions, and concrete recommendations for complexity limits, auth, and caching. I validate naming, nullable patterns, and federation composition to avoid N+1 and other common pitfalls.

When to use it

• Designing or evolving GraphQL schemas and type systems
• Implementing or troubleshooting Apollo Federation subgraphs
• Building resolvers optimized with DataLoader to avoid N+1
• Adding real-time subscriptions with WebSocket/pub-sub backends
• Improving query performance, complexity, and security

Best practices

• Follow schema-first design and document every type and field
• Use DataLoader for batching and caching in resolvers to prevent N+1
• Enforce query complexity and depth limits, plus persisted queries
• Apply field-level authorization and avoid hardcoded auth in resolvers
• Use federation directives correctly and keep service boundaries narrow

Example use cases

• Draft a federated schema for a multi-team e-commerce platform with product, inventory, and pricing subgraphs
• Convert a monolithic REST API to a schema-first GraphQL API with typed inputs and migrations
• Implement resolvers that batch DB calls for user and order joins and include DataLoader examples
• Design real-time order status subscriptions using Redis pub/sub or a scalable WebSocket layer
• Add query complexity scoring and sample persisted-query flow for a public GraphQL endpoint