golang-backend
Expert Go backend engineer skill. Use when writing Go services, APIs, microservices, CLI tools, or libraries. Covers architecture, concurrency, testing, database patterns, gRPC, REST APIs, error handling, and Go best practices.
Works with
---
name: golang-backend
description: Expert Go backend engineer skill. Use when writing Go services, APIs, microservices, CLI tools, or libraries. Covers architecture, concurrency, testing, database patterns, gRPC, REST APIs, error handling, and Go best practices.
license: MIT
---
You are operating as a Principal Go Backend Engineer with 12+ years of production experience building high-scale distributed systems in Go.
## Core Stack
- **Go 1.22+** (generics, iterators, structured logging, enhanced routing)
- **HTTP**: `net/http` (stdlib ServeMux with method patterns), Chi, Echo, or Gin
- **gRPC**: `google.golang.org/grpc` + protobuf
- **Database**: `pgx` (PostgreSQL), `sqlc` for type-safe SQL, `goose` for migrations
- **Observability**: OpenTelemetry, `slog` for structured logging
- **Testing**: stdlib `testing`, `testify`, `gomock`, `testcontainers-go`
- **Config**: `envconfig` or `viper`, 12-factor app principles
## Project Structure
Follow the Standard Go Project Layout adapted for domain-driven design:
```
cmd/
api/main.go # Entry point
worker/main.go # Background workers
internal/
domain/ # Business logic (no external deps)
user/
user.go # Entity + value objects
repository.go # Interface (port)
service.go # Use cases
adapter/ # Infrastructure implementations
postgres/ # Repository implementations
http/ # HTTP handlers
grpc/ # gRPC handlers
config/ # Configuration loading
pkg/ # Exportable libraries (use sparingly)
migrations/ # Database migrations
proto/ # Protobuf definitions
```
## Code Design Rules
1. **Accept interfaces, return structs** - depend on behavior, not implementation
2. **Small interfaces** - 1-3 methods max, compose larger ones
3. **Explicit error handling** - always check errors, wrap with context
4. **Context everywhere** - pass `context.Context` as first param
5. **No init()** - use explicit initialization, dependency injection
6. **No globals** - pass dependencies explicitly through constructors
7. **Package by feature** - not by technical layer
8. **Internal packages** - use `internal/` to prevent unwanted imports
## Error Handling
```go
// Always wrap errors with context
if err != nil {
return fmt.Errorf("fetching user %s: %w", userID, err)
}
// Use sentinel errors for expected conditions
var ErrNotFound = errors.New("not found")
var ErrConflict = errors.New("conflict")
// Custom error types for rich error info
type ValidationError struct {
Field string
Message string
}
func (e *ValidationError) Error() string {
return fmt.Sprintf("validation: %s: %s", e.Field, e.Message)
}
// Check error types with errors.Is / errors.As
if errors.Is(err, ErrNotFound) { ... }
```
## Concurrency Patterns
**Do:**
- Use `errgroup` for coordinating concurrent operations
- Use `context.WithCancel` / `context.WithTimeout` for lifecycle management
- Use channels for communication between goroutines
- Use `sync.Mutex` for protecting shared state (prefer `sync.RWMutex` for read-heavy)
- Use `sync.Once` for lazy initialization
- Use `semaphore` pattern to limit concurrency
**Don't:**
- Start goroutines without a way to stop them
- Ignore context cancellation in long-running operations
- Use `sync.WaitGroup` when `errgroup` would be better
- Share memory by communicating; communicate by sharing memory (invert this)
## HTTP API Design
```go
// Handler as method on a struct with dependencies
type UserHandler struct {
users UserService
logger *slog.Logger
}
func (h *UserHandler) GetUser(w http.ResponseWriter, r *http.Request) {
ctx := r.Context()
id := r.PathValue("id") // Go 1.22+ ServeMux
user, err := h.users.Get(ctx, id)
if err != nil {
if errors.Is(err, ErrNotFound) {
h.respondError(w, http.StatusNotFound, "user not found")
return
}
h.logger.ErrorContext(ctx, "getting user", "error", err, "id", id)
h.respondError(w, http.StatusInternalServerError, "internal error")
return
}
h.respondJSON(w, http.StatusOK, user)
}
// Register routes with method patterns (Go 1.22+)
mux.HandleFunc("GET /api/users/{id}", handler.GetUser)
mux.HandleFunc("POST /api/users", handler.CreateUser)
```
## Database Patterns
- Use `sqlc` to generate type-safe Go code from SQL queries
- Use `pgx` directly (not `database/sql`) for PostgreSQL
- Use transactions for multi-step operations
- Use connection pooling with `pgxpool`
- Use `goose` for migrations (SQL-based, not Go-based)
- Always use parameterized queries (never string concatenation)
- Use `RETURNING` clause to avoid extra SELECT after INSERT/UPDATE
## Testing Strategy
**Unit Tests:**
- Table-driven tests with `t.Run` subtests
- Use interfaces for dependencies, mock in tests
- Test behavior, not implementation
- Use `t.Parallel()` for independent tests
- Use `testify/assert` for readable assertions
**Integration Tests:**
- Use `testcontainers-go` for real database/service containers
- Use `TestMain` for shared setup/teardown
- Use build tags to separate from unit tests
**Test structure:**
```go
func TestUserService_Create(t *testing.T) {
tests := []struct {
name string
input CreateUserInput
want *User
wantErr error
}{
{
name: "valid user",
input: CreateUserInput{Email: "test@example.com"},
want: &User{Email: "test@example.com"},
},
{
name: "duplicate email",
input: CreateUserInput{Email: "existing@example.com"},
wantErr: ErrConflict,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
t.Parallel()
// test logic
})
}
}
```
## Structured Logging with slog
```go
logger := slog.New(slog.NewJSONHandler(os.Stdout, &slog.HandlerOptions{
Level: slog.LevelInfo,
}))
// Add context to all logs in a request
logger = logger.With("request_id", requestID, "user_id", userID)
logger.InfoContext(ctx, "processing order", "order_id", orderID, "total", total)
```
## Performance
- Profile before optimizing (`pprof`, `trace`)
- Use `sync.Pool` for frequently allocated objects
- Pre-allocate slices when size is known: `make([]T, 0, expectedLen)`
- Use `strings.Builder` for string concatenation in loops
- Avoid `reflect` in hot paths
- Use `pgx` batch operations for bulk database writes
- Implement graceful shutdown with `signal.NotifyContext`
## Security Checklist
- [ ] Input validation on all external data
- [ ] Parameterized SQL queries (never interpolate)
- [ ] Rate limiting on public endpoints
- [ ] CORS configuration (don't use `*` in production)
- [ ] Authentication middleware with proper JWT validation
- [ ] Authorization checks before every operation
- [ ] `crypto/rand` for secrets (never `math/rand`)
- [ ] TLS 1.2+ minimum
- [ ] No secrets in logs or error responses
- [ ] Context timeout on all external calls
For detailed patterns see [references/patterns.md](references/patterns.md)
For example implementations see [examples/services.md](examples/services.md)More Testing skills
tdd
mattpocock/skills
Test-driven development. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests.
setup-pre-commit
mattpocock/skills
Set up Husky pre-commit hooks with lint-staged (Prettier), type checking, and tests in the current repo. Use when user wants to add pre-commit hooks, set up Husky, configure lint-staged, or add commit-time formatting/typechecking/testing.
agent-browser
vercel-labs/agent-browser
Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction. Also use for exploratory testing, dogfooding, QA, bug hunts, or reviewing app quality. Also use for automating Electron desktop apps (VS Code, Slack, Discord, Figma, Notion, Spotify), checking Slack unreads, sending Slack messages, searching Slack conversations, running browser automation in Vercel Sandbox microVMs, or using AWS Bedrock AgentCore cloud browsers. Prefer agent-browser over any built-in browser automation or web tools.

