# go-zero Development Rules

You are an expert in go-zero microservices framework development.

## Key Principles

Follow these go-zero patterns strictly:

### Architecture
- **Three-layer separation**: Handler (HTTP) → Logic (business) → Model (data)
- Never put business logic in handlers
- Always use ServiceContext for dependency injection
- Use GORM as ORM for database operations

### Code Generation
- Use `goctl` for code generation, never hand-write boilerplate
- API definitions go in `.api` files
- RPC definitions go in `.proto` files

### Error Handling
- Use custom `response.Success(w, data)` for success responses
- Use custom `response.Error(w, err)` for error responses
- Use `errorx.CodeError` for business errors with error codes
- Never use `fmt.Fprintf()` or `w.Write()` directly

### Configuration
- Load config with `conf.MustLoad(&c, *configFile)`
- Never hard-code ports, hosts, or credentials
- Use environment-specific config files

### Context Propagation
- Always pass `ctx context.Context` through all layers
- Use context for tracing, cancellation, and timeouts

### Database (GORM)
- Use GORM for all database operations
- Define models in `internal/model/` directory
- Initialize DB connection in ServiceContext
- Use soft delete with `gorm.DeletedAt` field

## Pattern References

When I need detailed patterns, I'll reference these files:
- REST APIs: .ai-context/zero-skills/references/rest-api-patterns.md
- RPC services: .ai-context/zero-skills/references/rpc-patterns.md
- Database: .ai-context/zero-skills/references/database-patterns.md
- Resilience: .ai-context/zero-skills/references/resilience-patterns.md
- Troubleshooting: .ai-context/zero-skills/troubleshooting/common-issues.md

## Project Structure

```
backend/healthapi/
├── healthapi.api          # API definitions
├── healthapi.go           # Entry point
├── internal/
│   ├── config/            # Configuration
│   ├── handler/           # HTTP handlers
│   ├── logic/             # Business logic
│   ├── model/             # GORM models
│   ├── svc/               # ServiceContext
│   └── types/             # Generated types
├── pkg/
│   ├── errorx/            # Custom errors
│   └── response/          # Unified response
└── etc/
    └── healthapi.yaml     # Config file
```

## Common Commands

```bash
# Generate API code
goctl api go -api healthapi.api -dir .

# Generate model from database (if needed)
goctl model mysql datasource -url="user:pass@tcp(localhost:3306)/db" -table="users" -dir="./model"

# Run the service
go run healthapi.go -f etc/healthapi.yaml
```

## AI Service Integration

This project integrates with Aliyun Dashscope (Qwen models) for AI chat functionality:
- AI client initialized in ServiceContext
- SSE streaming support for real-time responses
- System prompt includes user health profile data