# ProtectMyAPI Integration Rules

You are helping integrate ProtectMyAPI into a mobile app. ProtectMyAPI protects API keys by proxying requests through a secure server with device attestation.

## Key Concepts

- API keys are stored in the ProtectMyAPI dashboard, NOT in the app
- The SDK verifies the device is legitimate before forwarding requests
- Works with iOS (App Attest), Android (Play Integrity), and Flutter

## Quick Reference

### iOS Setup
```swift
import ProtectMyAPI

// In App init or AppDelegate
ProtectMyAPI.shared.configure(
    appId: "APP_ID_FROM_DASHBOARD",
    environment: .production
)

// Use AI services
let openai = ProtectMyAPI.openAIService()
let response = try await openai.chat(message: "Hello", model: "gpt-4o-mini")
```

### Android Setup
```kotlin
import com.protectmyapi.ProtectMyAPI

// In Application.onCreate()
ProtectMyAPI.initialize(
    context = this,
    appId = "APP_ID_FROM_DASHBOARD",
    environment = Environment.PRODUCTION
)

// Use AI services
val openai = ProtectMyAPIAI.openAIService()
val response = openai.chat(message = "Hello", model = "gpt-4o-mini")
```

### Flutter Setup
```dart
import 'package:protectmyapi/protectmyapi.dart';

// In main()
await ProtectMyAPI.initialize(
  appId: 'APP_ID_FROM_DASHBOARD',
  environment: Environment.production,
);

// Use AI services
final openai = ProtectMyAPIAI.openAIService();
final response = await openai.chat(message: "Hello", model: "gpt-4o-mini");
```

## Available Services

```
ProtectMyAPI.openAIService()      - ChatGPT, DALL-E, Whisper
ProtectMyAPI.anthropicService()   - Claude
ProtectMyAPI.geminiService()      - Google Gemini
ProtectMyAPI.mistralService()     - Mistral AI
ProtectMyAPI.groqService()        - Groq (fast inference)
ProtectMyAPI.deepSeekService()    - DeepSeek (reasoning)
ProtectMyAPI.stabilityService()   - Image generation
ProtectMyAPI.elevenLabsService()  - Text-to-speech
ProtectMyAPI.perplexityService()  - AI search with citations
ProtectMyAPI.togetherService()    - Open source models
ProtectMyAPI.replicateService()   - ML models
ProtectMyAPI.fireworksService()   - Fast LLMs
ProtectMyAPI.openRouterService()  - Multi-provider routing
ProtectMyAPI.braveSearchService() - Web search
ProtectMyAPI.deepLService()       - Translation
ProtectMyAPI.falService()         - Fast image gen
ProtectMyAPI.openMeteoService()   - Weather (free)
```

## Common Patterns

### Chat with OpenAI
```swift
let openai = ProtectMyAPI.openAIService()
let response = try await openai.createChatCompletion(
    request: OpenAIChatRequest(
        model: "gpt-4o",
        messages: [
            .system("You are helpful."),
            .user("Hello!")
        ]
    )
)
print(response.choices.first?.message.content ?? "")
```

### Streaming
```swift
for try await chunk in openai.createChatCompletionStream(request: request) {
    print(chunk.choices.first?.delta?.content ?? "", terminator: "")
}
```

### Image Generation
```swift
let stability = ProtectMyAPI.stabilityService()
let image = try await stability.generateImageUltra(
    prompt: "A sunset over mountains",
    options: StabilityUltraOptions(aspectRatio: "16:9")
)
```

### Text-to-Speech
```swift
let elevenlabs = ProtectMyAPI.elevenLabsService()
let audio = try await elevenlabs.textToSpeech(
    text: "Hello world",
    voiceId: "21m00Tcm4TlvDq8ikWAM"
)
```

## Important Rules

1. NEVER put API keys in app code - they go in the dashboard
2. Use .development environment for testing (allows simulators)
3. Use .production environment for release builds
4. Attestation requires real devices, not simulators
5. Always handle errors (attestation can fail)

## Error Handling
```swift
do {
    let response = try await openai.chat(message: "Hi", model: "gpt-4o")
} catch ProtectMyAPIError.attestationFailed {
    // Device failed verification (simulator, jailbroken, etc.)
} catch ProtectMyAPIError.networkError {
    // Network issue
} catch ProtectMyAPIError.rateLimited {
    // Too many requests
}
```

## Full Documentation

For complete API reference: https://docs.protectmyapi.com
For comprehensive AI context: https://docs.protectmyapi.com/llms-full.txt