golang-graphql

by samber

Implements GraphQL APIs in Golang using gqlgen or graphql-go. Apply when building GraphQL servers, designing schemas, writing resolvers, handling subscriptions, or integrating GraphQL with existing Go HTTP services. Also apply when the codebase imports github.com/99designs/gqlgen or github.com/graph-gophers/graphql-go. npx skills add https://github.com/samber/cc-skills-golang --skill golang-graphql Download ZIPGitHub2.4k Persona: You are a Go GraphQL engineer. You design schemas deliberately, batch database access to prevent N+1, and treat query complexity limits as non-optional in production.

Modes:

• Build mode — generating new schemas, resolvers, or server setup: follow the skill's sequential instructions; launch a background agent to grep for existing resolver patterns and naming conventions before generating new code.

• Review mode — auditing a GraphQL codebase or PR: use a sub-agent to scan for N+1 resolver patterns, missing complexity caps, global DataLoaders, and introspection enabled in production, in parallel with reading the business logic.

Community default. A company skill that explicitly supersedes samber/cc-skills-golang@golang-graphql skill takes precedence.

Go GraphQL Best Practices

Both major libraries are schema-first: write SDL (.graphql files), bind Go resolvers. Choose based on project size and team preferences.

This skill is not exhaustive. Refer to each library's official documentation and code examples for current API signatures. Context7 can help as a discoverability platform. For Go package docs, versions, symbols, and known vulnerabilities, → See samber/cc-skills-golang@golang-pkg-go-dev skill.

Library Choice

Library Approach Type safety Build step Best for github.com/99designs/gqlgen Codegen Compile-time go generate Large schemas, federation, strict types github.com/graph-gophers/graphql-go Reflection Parse-time None Simple schemas, fast iteration github.com/graphql-go/graphql Code-first Runtime None Avoid — verbose, no SDL

Pick gqlgen when: Apollo Federation is required, schema is large (100+ types), or the team wants generated stubs and zero reflection overhead.

Pick graph-gophers when: schema is small/medium, the build pipeline should stay simple, or a dynamic schema is needed.

For deep-dive on each library, see gqlgen reference and graphql-go reference.

Schema Design

# ✓ Good — explicit nullability; ID scalar for opaque identifiers
type User {
 id: ID!
 email: String! # non-null: the server can always return this
 bio: String # nullable: may be unset
 posts(first: Int = 10, after: String): PostConnection!
}

# ✗ Bad — Int ID leaks implementation details, breaks client caching
type Post {
 id: Int!
}

Nullability rule: mark a field ! only when the server can always return a value. A resolver error on a non-null field nulls the parent object, causing cascade failures; nullable fields only null the field itself.

Pagination: use Relay cursor connections (Connection/Edge/PageInfo) for list fields. Avoid offset pagination on large datasets — cursors are stable under concurrent writes.

Mutations: wrap results in an envelope type so clients receive business errors alongside partial results without polluting the GraphQL errors array:

type CreateUserPayload {
 user: User
 errors: [UserError!]!
}

Resolver Patterns

Keep resolvers thin — they translate GraphQL inputs to domain calls and domain responses to GraphQL outputs.

// ✓ Good — resolver delegates to service layer
func (r *mutationResolver) CreateUser(ctx context.Context, input model.CreateUserInput) (*model.CreateUserPayload, error) {
 user, err := r.userService.Create(ctx, input.Email, input.Name)
 if err != nil {
 return nil, formatError(err)
 }
 return &model.CreateUserPayload{User: toGQLUser(user)}, nil
}

// ✗ Bad — SQL in resolver, no separation of concerns
func (r *queryResolver) User(ctx context.Context, id string) (*model.User, error) {
 row := r.db.QueryRowContext(ctx, "SELECT * FROM users WHERE id = $1", id)
 // ...
}

Use per-type resolver structs (userResolver, postResolver) rather than one monolithic resolver for all fields.

N+1 Prevention (DataLoaders)

Each User.posts resolver fires a SQL query per user without batching — O(n) DB calls for n users. DataLoaders solve this by coalescing per-field loads into a single batch query.

Critical rule: DataLoaders MUST be created per-request in HTTP middleware, never globally. A global DataLoader caches across requests — stale data, potential cross-user data leakage.

// ✓ Good — per-request DataLoader in middleware
func DataLoaderMiddleware(db *sql.DB, next http.Handler) http.Handler {
 return http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
 loaders := &Loaders{
 PostsByUserID: newPostsByUserIDLoader(r.Context(), db),
 }
 ctx := context.WithValue(r.Context(), loadersKey, loaders)
 next.ServeHTTP(w, r.WithContext(ctx))
 })
}

// ✗ Bad — global DataLoader shared across all requests
var globalLoader = newPostsByUserIDLoader(context.Background(), db)

In gqlgen, mark batched fields with resolver: true in gqlgen.yml to force a dedicated resolver method. See gqlgen reference for full DataLoader wiring.

Authentication and Authorization

Two-layer model:

• HTTP middleware — extract and validate tokens, stash identity in context.Context.

• Schema directives (gqlgen) or resolver checks (graphql-go) — enforce per-field authorization.

// HTTP middleware layer (both libraries)
func AuthMiddleware(next http.Handler) http.Handler {
 return http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
 token := r.Header.Get("Authorization")
 user, err := validateToken(token)
 if err != nil {
 http.Error(w, "Unauthorized", http.StatusUnauthorized)
 return
 }
 ctx := context.WithValue(r.Context(), userKey, user)
 next.ServeHTTP(w, r.WithContext(ctx))
 })
}

In gqlgen, use @hasRole schema directives for field-level authorization — authorization policy lives in the schema, not scattered across resolvers. See gqlgen reference.

Error Handling

Never return raw internal errors — they leak SQL messages, stack traces, or service internals to clients.

// gqlgen — custom ErrorPresenter strips internal details
srv.SetErrorPresenter(func(ctx context.Context, err error) *gqlerror.Error {
 var gqlErr *gqlerror.Error
 if errors.As(err, &gqlErr) {
 return gqlErr // already formatted
 }
 // log internal err here
 return gqlerror.Errorf("internal error") // safe client message
})

// Add extension codes for client-side error handling
return nil, &gqlerror.Error{
 Message: "user not found",
 Extensions: map[string]any{"code": "NOT_FOUND"},
}

For graph-gophers, implement the ResolverError interface to attach Extensions(). See graphql-go reference.

Use graphql.AddError(ctx, err) in gqlgen for non-fatal field errors where the resolver can still return partial data.

For error wrapping patterns, see the samber/cc-skills-golang@golang-error-handling skill.

Subscriptions

Subscriptions use long-lived WebSocket connections. The critical discipline: always respect context cancellation — a leaked goroutine per disconnected client exhausts resources silently.

// ✓ Good — closes channel when client disconnects
func (r *subscriptionResolver) MessageAdded(ctx context.Context, room string) ( Production GraphQL servers require explicit limits. Without them, a single deeply nested query exhausts CPU and memory.

// gqlgen — wire these into every production handler srv := handler.NewDefaultServer(es) srv.Use(extension.FixedComplexityLimit(200)) // max cost per query

// Gate introspection — only in non-production environments if os.Getenv("ENV") != "production" { srv.Use(extension.Introspection{}) }

 For graph-gophers: `graphql.MaxDepth(10)` and `graphql.MaxParallelism(10)` options at `ParseSchema` time.

 **Query allow-listing:** in production, consider persisted queries (gqlgen APQ extension) to reject arbitrary query strings.

## Deep Dives

- **[gqlgen reference](https://github.com/samber/cc-skills-golang/blob/main/skills/golang-graphql/references/gqlgen.md)** — codegen workflow, `gqlgen.yml`, DataLoaders, Federation v2, directives 

- **[graphql-go reference](https://github.com/samber/cc-skills-golang/blob/main/skills/golang-graphql/references/graphql-go.md)** — reflection resolver model, type mapping, tracing 

- **[Testing](https://github.com/samber/cc-skills-golang/blob/main/skills/golang-graphql/references/testing.md)** — gqlgen client harness, gqltesting, httptest patterns

## Cross-References

- → See `samber/cc-skills-golang@golang-context` skill for context propagation in resolvers and subscriptions 

- → See `samber/cc-skills-golang@golang-error-handling` skill for error wrapping and sentinel patterns 

- → See `samber/cc-skills-golang@golang-testing` skill for table-driven and integration test patterns 

- → See `samber/cc-skills-golang@golang-observability` skill for tracing and metrics in resolvers 

- → See `samber/cc-skills-golang@golang-security` skill for input validation and injection prevention 

- → See `samber/cc-skills-golang@golang-database` skill for N+1 query patterns and DataLoader database batching

## References

- [gqlgen](https://github.com/99designs/gqlgen) 

- [graph-gophers/graphql-go](https://github.com/graph-gophers/graphql-go) 

- [Relay cursor connections spec](https://relay.dev/graphql/connections.htm) 

 If you encounter a bug or unexpected behavior in gqlgen, open an issue at [https://github.com/99designs/gqlgen/issues](https://github.com/99designs/gqlgen/issues).

 If you encounter a bug or unexpected behavior in graph-gophers/graphql-go, open an issue at [https://github.com/graph-gophers/graphql-go/issues](https://github.com/graph-gophers/graphql-go/issues).