### Custom Welcome Message Example

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/ChatViewModel.md

Example of creating a custom welcome message dynamically and passing it to LlmChatView.

```dart
class ChatPage extends StatefulWidget {
  @override
  State<ChatPage> createState() => _ChatPageState();
}

class _ChatPageState extends State<ChatPage> {
  late String welcomeMessage;

  @override
  void initState() {
    super.initState();
    final userName = 'Alice';
    welcomeMessage = 'Hello $userName! Welcome to the AI Chat Assistant.';
  }

  @override
  Widget build(BuildContext context) {
    return LlmChatView(
      provider: FirebaseProvider(model: model),
      welcomeMessage: welcomeMessage,
      suggestions: [
        'What is my account status?',
        'Help with my order',
        'Chat with support',
      ],
      responseBuilder: (context, response) {
        return Column(
          crossAxisAlignment: CrossAxisAlignment.start,
          children: [
            SelectableText(response),
            SizedBox(height: 8),
            ActionChip(
              label: const Text('Copy'),
              onPressed: () {
                Clipboard.setData(ClipboardData(text: response));
              },
            ),
          ],
        );
      },
    );
  }
}
```

--------------------------------

### Complete LlmChatView Configuration Example

Source: https://github.com/flutter/ai/blob/main/_autodocs/configuration.md

This example demonstrates a full configuration of the LlmChatView widget, including custom styling, strings, suggestions, and error handling. It shows how to integrate with Firebase for AI model interaction.

```dart
void main() async {
  WidgetsFlutterBinding.ensureInitialized();
  await Firebase.initializeApp();
  runApp(MyApp());
}

class MyApp extends StatelessWidget {
  @override
  Widget build(BuildContext context) {
    final customStyle = LlmChatViewStyle(
      backgroundColor: Color(0xFFF5F5F5),
      messageSpacing: 12.0,
      userMessageStyle: UserMessageStyle(
        textStyle: TextStyle(fontSize: 16, color: Colors.black87),
        decoration: BoxDecoration(
          color: Color(0xFFE3F2FD),
          borderRadius: BorderRadius.circular(16),
        ),
      ),
      llmMessageStyle: LlmMessageStyle(
        decoration: BoxDecoration(
          color: Color(0xFFEEEEEE),
          borderRadius: BorderRadius.circular(16),
        ),
        padding: EdgeInsets.all(12),
      ),
      chatInputStyle: ChatInputStyle(
        hintText: 'Ask me anything...',
        decoration: BoxDecoration(
          border: Border.all(color: Color(0xFFCCCCCC)),
          borderRadius: BorderRadius.circular(24),
        ),
      ),
    );

    final customStrings = LlmChatViewStrings(
      send: 'Send',
      typeAMessage: 'Your message here...',
      // ... other strings
    );

    return MaterialApp(
      home: Scaffold(
        appBar: AppBar(title: Text('AI Chat')),
        body: LlmChatView(
          provider: FirebaseProvider(
            model: FirebaseAI.vertexAI().generativeModel(
              model: 'gemini-2.0-flash',
            ),
            chatGenerationConfig: GenerationConfig(temperature: 0.7),
          ),
          style: customStyle,
          strings: customStrings,
          suggestions: ['Hello', 'Help me with something'],
          welcomeMessage: 'Welcome to AI Chat!',
          enableAttachments: true,
          enableVoiceNotes: true,
          onErrorCallback: (context, error) {
            ScaffoldMessenger.of(context).showSnackBar(
              SnackBar(content: Text('Error: ${error.message}')),
            );
          },
        ),
      ),
    );
  }
}
```

--------------------------------

### Suggestions Example

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/ChatViewModel.md

Provides predefined suggestions to guide users when the chat is empty. This enhances user experience by offering quick-start prompts.

```dart
suggestions: [
  'What is Flutter?',
  'How do I use widgets?',
  'Tell me a joke',
]
```

--------------------------------

### Default ActionButtonStyle Examples

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/ActionButtonStyle.md

Examples demonstrating how to create default styles for submit and record button types.

```dart
final sendStyle = ActionButtonStyle.defaultStyle(
  ActionButtonType.submit,
  strings: customStrings,
);

final recordStyle = ActionButtonStyle.defaultStyle(
  ActionButtonType.record,
);
```

--------------------------------

### Simple LlmChatView Initialization

Source: https://github.com/flutter/ai/blob/main/_autodocs/README.md

Basic setup for the LlmChatView widget with a required FirebaseProvider.

```dart
return Scaffold(
  body: LlmChatView(
    provider: FirebaseProvider(model: model),
  ),
);
```

--------------------------------

### Flutter App Setup with FirebaseProvider

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/FirebaseProvider.md

Initialize Firebase and set up FirebaseProvider within a Flutter application. This example demonstrates integrating FirebaseAI with LlmChatView for a chat interface.

```dart
import 'package:firebase_core/firebase_core.dart';
import 'package:firebase_ai/firebase_ai.dart';
import 'package:flutter_ai_toolkit/flutter_ai_toolkit.dart';

void main() async {
  WidgetsFlutterBinding.ensureInitialized();
  await Firebase.initializeApp();
  runApp(const MyApp());
}

class ChatPage extends StatelessWidget {
  @override
  Widget build(BuildContext context) {
    final provider = FirebaseProvider(
      model: FirebaseAI.vertexAI().generativeModel(
        model: 'gemini-2.0-flash',
      ),
    );

    return Scaffold(
      appBar: AppBar(title: const Text('Firebase AI Chat')),
      body: LlmChatView(
        provider: provider,
        suggestions: ['Hello', 'How are you?'],
        enableAttachments: true,
      ),
    );
  }
}
```

--------------------------------

### Example of copyWith() Usage

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmChatViewStyle.md

Demonstrates how to create a new LlmChatViewStyle by copying an existing one and overriding specific properties.

```dart
final baseStyle = LlmChatViewStyle.defaultStyle();
final customStyle = baseStyle.copyWith(
  backgroundColor: Colors.black,
  messageSpacing: 16.0,
  padding: const EdgeInsets.all(16),
);
```

--------------------------------

### Welcome Message Example

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/ChatViewModel.md

An optional welcome message to greet users when they first open the chat interface. It can be customized to provide initial guidance.

```dart
welcomeMessage: 'Welcome to the AI Assistant! How can I help you today?'
```

--------------------------------

### LinkAttachment Creation Example

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/Attachment.md

Demonstrates how to create LinkAttachment instances, showing both inferred and explicitly set MIME types.

```dart
final link = LinkAttachment(
  name: 'Flutter Documentation',
  url: Uri.parse('https://flutter.dev/docs'),
); // MIME type inferred as 'text/html'

final pdfLink = LinkAttachment(
  name: 'Flutter Guide',
  url: Uri.parse('https://example.com/guide.pdf'),
  mimeType: 'application/pdf', // Explicitly set
);
```

--------------------------------

### Dart Code Block Example

Source: https://github.com/flutter/ai/blob/main/_autodocs/MANIFEST.md

Illustrates the standard Markdown formatting for Dart code examples within the documentation.

```dart
// Dart code example
```

--------------------------------

### Resolve ActionButtonStyle Example

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/ActionButtonStyle.md

Example showing how to resolve a style, overriding the icon color of a default stop button.

```dart
final resolved = ActionButtonStyle.resolve(
  ActionButtonStyle(iconColor: Colors.red),
  defaultStyle: ActionButtonStyle.defaultStyle(ActionButtonType.stop),
);
// Result: Stop button style with custom red color
```

--------------------------------

### Full ChatMessage Usage Example

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/ChatMessage.md

Demonstrates creating a conversation, handling user messages, simulating LLM streaming responses, and serializing/deserializing the conversation.

```dart
import 'package:flutter_ai_toolkit/flutter_ai_toolkit.dart';

void main() async {
  // Create a simple conversation
  List<ChatMessage> history = [];

  // User sends message
  history.add(ChatMessage.user('What is Flutter?', []));

  // LLM responds with streaming
  final response = ChatMessage.llm();
  history.add(response);

  // Simulate streaming response
  final stream = provider.generateStream('What is Flutter?');
  await for (final chunk in stream) {
    response.append(chunk);
    print('Current response: ${response.text}');
  }

  // Serialize conversation
  final savedConversation = history.map((m) => m.toJson()).toList();
  await storage.saveMessages(savedConversation);

  // Restore conversation
  final loaded = (await storage.loadMessages())
    .map(ChatMessage.fromJson)
    .toList();
  provider.history = loaded;
}
```

--------------------------------

### generateStream Example Usage

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmProvider.md

Demonstrates how to call generateStream and handle the resulting text stream, including error and completion callbacks.

```dart
final stream = provider.generateStream(
  'Translate this image to text',
  attachments: [imageAttachment],
);

stream.listen(
  (chunk) => print('Received: $chunk'),
  onError: (error) => print('Error: $error'),
  onDone: () => print('Complete'),
);
```

--------------------------------

### Dark Theme Customization

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmChatViewStyle.md

Provides an example of how to create a dark theme for the `LlmChatView` by setting appropriate background and message colors.

```APIDOC
## Dark Theme Customization

### Description
Applies a dark theme to the `LlmChatView` by adjusting background and message container colors.

### Parameters
- **backgroundColor** (Color) - The dark background color for the chat view.
- **userMessageStyle.decoration.color** (Color) - The dark background color for user message bubbles.
- **llmMessageStyle.decoration.color** (Color) - The dark background color for LLM message bubbles.

### Example
```dart
final darkStyle = LlmChatViewStyle(
  backgroundColor: Color(0xFF121212),
  userMessageStyle: UserMessageStyle(
    decoration: BoxDecoration(color: Color(0xFF1E1E1E), borderRadius: BorderRadius.circular(12)),
  ),
  llmMessageStyle: LlmMessageStyle(
    decoration: BoxDecoration(color: Color(0xFF2C2C2C), borderRadius: BorderRadius.circular(12)),
  ),
);
```
```

--------------------------------

### Basic LlmChatView Configuration

Source: https://github.com/flutter/ai/blob/main/_autodocs/README.md

Configure the LlmChatView with a basic FirebaseProvider. This is the minimal setup required to display a chat interface.

```dart
LlmChatView(
  provider: FirebaseProvider(model: model),
)
```

--------------------------------

### Custom Styled Chat Buttons Example

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/ActionButtonStyle.md

Demonstrates how to customize individual action button styles within the LlmChatView.

```dart
import 'package:flutter/material.dart';
import 'package:flutter_ai_toolkit/flutter_ai_toolkit.dart';

class CustomStyledChat extends StatelessWidget {
  @override
  Widget build(BuildContext context) {
    final customStyle = LlmChatViewStyle(
      // Customize individual button styles
      submitButtonStyle: ActionButtonStyle(
        icon: Icons.send,
        iconColor: Colors.blue,
        iconDecoration: BoxDecoration(
          color: Colors.blue.shade50,
          shape: BoxShape.circle,
          boxShadow: [
            BoxShadow(
              color: Colors.blue.withOpacity(0.3),
              blurRadius: 8,
            ),
          ],
        ),
        text: 'Send Message',
        textStyle: TextStyle(
          fontSize: 12,
          color: Colors.blue,
          fontWeight: FontWeight.w500,
        ),
      ),
      
      recordButtonStyle: ActionButtonStyle(
        icon: Icons.mic,
        iconColor: Colors.red,
        text: 'Record',
      ),
      
      stopButtonStyle: ActionButtonStyle(
        icon: Icons.stop_circle,
        iconColor: Colors.red,
        iconDecoration: BoxDecoration(
          color: Colors.red.shade50,
          shape: BoxShape.circle,
        ),
      ),
      
      attachFileButtonStyle: ActionButtonStyle(
        icon: Icons.attach_file,
        iconColor: Colors.green,
        text: 'Attach',
      ),
    );

    return Scaffold(
      appBar: AppBar(title: const Text('Custom Buttons')),
      body: LlmChatView(
        provider: myProvider,
        style: customStyle,
      ),
    );
  }
}
```

--------------------------------

### Usage Example: Creating and Sending Attachments

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/Attachment.md

Shows how to create various attachment types (image, file, link) and send them in a chat message using the provider.

```dart
import 'package:flutter_ai_toolkit/flutter_ai_toolkit.dart';
import 'package:image_picker/image_picker.dart';
import 'package:file_selector/file_selector.dart';

Future<void> attachmentExample() async {
  final ImagePicker picker = ImagePicker();
  final attachments = <Attachment>[];

  // Attach an image
  final image = await picker.pickImage(source: ImageSource.gallery);
  if (image != null) {
    final imageAttachment = await ImageFileAttachment.fromFile(image);
    attachments.add(imageAttachment);
  }

  // Attach a PDF file
  const xFileFilter = XTypeGroup(
    label: 'PDFs',
    extensions: ['pdf'],
  );
  final file = await openFile(acceptedTypeGroups: [xFileFilter]);
  if (file != null) {
    final fileAttachment = await FileAttachment.fromFile(file);
    attachments.add(fileAttachment);
  }

  // Attach a link
  attachments.add(LinkAttachment(
    name: 'Reference',
    url: Uri.parse('https://api.example.com/docs'),
  ));

  // Send message with attachments
  await provider.sendMessageStream(
    'Please analyze these files',
    attachments: attachments,
  ).forEach(print);
```

--------------------------------

### Basic LlmChatView Usage

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmChatView.md

This snippet shows the basic setup for LlmChatView, including Firebase initialization and configuration with a generative model. It demonstrates how to set up the chat interface with suggestions, a welcome message, and attachment/voice note capabilities.

```dart
import 'package:flutter/material.dart';
import 'package:firebase_ai/firebase_ai.dart';
import 'package:firebase_core/firebase_core.dart';
import 'package:flutter_ai_toolkit/flutter_ai_toolkit.dart';

void main() async {
  WidgetsFlutterBinding.ensureInitialized();
  await Firebase.initializeApp();
  runApp(const MyApp());
}

class MyApp extends StatelessWidget {
  const MyApp({Key? key}) : super(key: key);

  @override
  Widget build(BuildContext context) {
    return MaterialApp(
      home: ChatPage(),
    );
  }
}

class ChatPage extends StatelessWidget {
  @override
  Widget build(BuildContext context) {
    return Scaffold(
      appBar: AppBar(title: const Text('AI Chat')),
      body: LlmChatView(
        provider: FirebaseProvider(
          model: FirebaseAI.vertexAI().generativeModel(
            model: 'gemini-2.0-flash',
          ),
        ),
        style: LlmChatViewStyle(
          backgroundColor: Colors.white,
        ),
        suggestions: [
          'What is Flutter?',
          'How do I use Firebase?',
          'Tell me a joke',
        ],
        welcomeMessage: 'Welcome! Ask me anything.',
        enableAttachments: true,
        enableVoiceNotes: true,
      ),
    );
  }
}
```

--------------------------------

### Example of Resolving LlmChatViewStyle

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmChatViewStyle.md

Demonstrates how to create a custom style by overriding specific properties and falling back to default styles for others.

```dart
final resolved = LlmChatViewStyle.resolve(
  LlmChatViewStyle(backgroundColor: Colors.white),
  defaultStyle: LlmChatViewStyle.defaultStyle(),
);
// Result: Custom background with default styling for all other properties
```

--------------------------------

### Basic LlmChatView Styling

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmChatViewStyle.md

Provides a comprehensive example of styling LlmChatView with custom colors, message styles, input styles, and spacing.

```dart
import 'package:flutter/material.dart';
import 'package:flutter_ai_toolkit/flutter_ai_toolkit.dart';

class ChatPage extends StatelessWidget {
  @override
  Widget build(BuildContext context) {
    return Scaffold(
      appBar: AppBar(title: const Text('Custom Styled Chat')),
      body: LlmChatView(
        provider: myProvider,
        style: LlmChatViewStyle(
          // Background
          backgroundColor: Color(0xFFF5F5F5),
          menuColor: Colors.white,
          progressIndicatorColor: Colors.blue,

          // Message styling
          userMessageStyle: UserMessageStyle(
            textStyle: TextStyle(fontSize: 16),
            decoration: BoxDecoration(
              color: Colors.blue.shade100,
              borderRadius: BorderRadius.circular(12),
            ),
          ),
          llmMessageStyle: LlmMessageStyle(
            textStyle: TextStyle(fontSize: 16),
            decoration: BoxDecoration(
              color: Colors.grey.shade200,
              borderRadius: BorderRadius.circular(12),
            ),
          ),

          // Input styling
          chatInputStyle: ChatInputStyle(
            hintText: 'Type your message...',
            backgroundColor: Colors.white,
            decoration: BoxDecoration(
              border: Border.all(color: Colors.grey),
              borderRadius: BorderRadius.circular(8),
            ),
          ),

          // Spacing
          messageSpacing: 12.0,
          padding: EdgeInsets.all(16),
          margin: EdgeInsets.symmetric(vertical: 8),

          // Suggestions
          suggestionStyle: SuggestionStyle(
            textStyle: TextStyle(fontSize: 14),
            decoration: BoxDecoration(
              color: Colors.blue.shade50,
              borderRadius: BorderRadius.circular(20),
              border: Border.all(color: Colors.blue),
            ),
          ),
        ),
      ),
    );
  }
}
```

--------------------------------

### MessageOrigin Getters Example

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/MessageOrigin.md

Demonstrates how to use the `isUser` and `isLlm` getters to check the origin of a message.

```dart
final origin = MessageOrigin.user;
print(origin.isUser);   // true
print(origin.isLlm);    // false
```

--------------------------------

### sendMessageStream Example Usage

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmProvider.md

Shows how to use sendMessageStream to get a response based on the current conversation and verifies that the history is updated.

```dart
final stream = provider.sendMessageStream('What was I asking before?');

await stream.forEach((chunk) {
  print('Streaming: $chunk');
});

// History now contains the user message and full LLM response
print('Chat history: ${provider.history}');
```

--------------------------------

### EchoProvider Usage Example

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/EchoProvider.md

Demonstrates how to use EchoProvider to generate streams of messages, both with and without file attachments. This is useful for testing chat interfaces and message handling.

```dart
import 'package:flutter_ai_toolkit/flutter_ai_toolkit.dart';
import 'dart:io';
import 'dart:typed_data';

void testEchoProvider() {
  final provider = EchoProvider();

  // Test simple echo
  final stream = provider.generateStream('Test message');
  stream.listen(print);

  // Test with attachments
  final file = FileAttachment(
    name: 'document.txt',
    mimeType: 'text/plain',
    bytes: Uint8List.fromList([1, 2, 3]),
  );

  final attachmentStream = provider.generateStream(
    'Document contents',
    [file],
  );
  attachmentStream.listen(print);
}

void testWithLlmChatView() {
  // Use EchoProvider for testing UI without Firebase
  return Scaffold(
    body: LlmChatView(
      provider: EchoProvider(),
      suggestions: ['Hello', 'Test'],
    ),
  );
}
```

--------------------------------

### LlmChatView with Suggestions

Source: https://github.com/flutter/ai/blob/main/_autodocs/README.md

Configure the LlmChatView to display suggestion chips to the user. These suggestions can guide the user's interaction with the AI.

```dart
LlmChatView(
  provider: provider,
  suggestions: ['What is Flutter?', 'How do I build widgets?'],
)
```

--------------------------------

### Response Builder Example

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/ChatViewModel.md

Implement a custom ResponseBuilder to control how LLM responses are rendered. This allows for advanced formatting like syntax highlighting or markdown.

```dart
responseBuilder: (context, response) {
  return SelectableText(
    response,
    style: Theme.of(context).textTheme.bodyMedium,
  );
}
```

--------------------------------

### Building a Message Display with Alignment and Styling

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/MessageOrigin.md

An example of a `StatelessWidget` that displays a list of messages, aligning them to the right for user messages and left for LLM messages, and applying different bubble colors.

```dart
class ConversationView extends StatelessWidget {
  final LlmProvider provider;

  @override
  Widget build(BuildContext context) {
    return ListView.builder(
      itemCount: provider.history.length,
      itemBuilder: (context, index) {
        final message = provider.history.elementAt(index);

        return Container(
          alignment: message.origin.isUser
              ? Alignment.centerRight
              : Alignment.centerLeft,
          padding: EdgeInsets.all(8),
          child: Bubble(
            color: message.origin.isUser ? Colors.blue : Colors.grey,
            child: Text(message.text ?? ''),
          ),
        );
      },
    );
  }
}
```

--------------------------------

### Spanish Localization Example

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmChatViewStrings.md

Provides a Spanish localization for LlmChatViewStrings. Use this when the application's locale is Spanish.

```dart
const spanishStrings = LlmChatViewStrings(
  addAttachment: 'Agregar archivo',
  attachFile: 'Adjuntar archivo',
  takePhoto: 'Tomar foto',
  stop: 'Parar',
  close: 'Cerrar',
  cancel: 'Cancelar',
  copyToClipboard: 'Copiar al portapapeles',
  editMessage: 'Editar mensaje',
  attachImage: 'Adjuntar imagen',
  recordAudio: 'Grabar audio',
  submitMessage: 'Enviar mensaje',
  closeMenu: 'Cerrar menú',
  unableToRecordAudio: 'No se puede grabar audio',
  unsupportedImageSource: 'Fuente de imagen no compatible',
  unableToPickImage: 'No se puede seleccionar imagen',
  unableToPickFile: 'No se puede seleccionar archivo',
  unableToPickUrl: 'No se puede seleccionar URL',
  messageCopiedToClipboard: 'Mensaje copiado al portapapeles',
  editing: 'Editando',
  error: 'Error',
  cancelMessage: 'Cancelar',
  submit: 'Enviar',
  send: 'Enviar',
  typeAMessage: 'Escribe un mensaje...', 
  recording: 'Grabando',
  tapToStop: 'Toca para parar',
  tapToRecord: 'Toca para grabar',
  releaseToCancel: 'Suelta para cancelar',
);

LlmChatView(
  provider: myProvider,
  strings: spanishStrings,
)
```

--------------------------------

### ChatMessage.llm Factory Constructor and Streaming Example

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/ChatMessage.md

Creates an LLM-originated message with an empty initial state, suitable for streaming responses. Use the `append()` method to add text as it arrives.

```dart
factory ChatMessage.llm()
```

```dart
final response = ChatMessage.llm();
provider.history.add(response);

// As response streams in:
stream.listen((chunk) {
  response.append(chunk);
});
```

--------------------------------

### Custom LLM Message Styling Example

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmMessageStyle.md

Demonstrates how to apply custom styling to LLM messages within a LlmChatView, including icon, decoration, and markdown options.

```dart
import 'package:flutter/material.dart';
import 'package:flutter_markdown_plus/flutter_markdown_plus.dart';
import 'package:flutter_ai_toolkit/flutter_ai_toolkit.dart';

class CustomChatView extends StatelessWidget {
  @override
  Widget build(BuildContext context) {
    return LlmChatView(
      provider: myProvider,
      style: LlmChatViewStyle(
        llmMessageStyle: LlmMessageStyle(
          icon: Icons.smart_toy,
          iconColor: Colors.deepPurple,
          iconDecoration: BoxDecoration(
            color: Colors.deepPurple.shade50,
            shape: BoxShape.circle,
          ),
          decoration: BoxDecoration(
            color: Colors.grey.shade100,
            borderRadius: BorderRadius.only(
              topLeft: Radius.circular(4),
              topRight: Radius.circular(20),
              bottomLeft: Radius.circular(20),
              bottomRight: Radius.circular(20),
            ),
            border: Border.all(
              color: Colors.grey.shade300,
              width: 1,
            ),
          ),
          markdownStyle: MarkdownStyleData(
            p: TextStyle(fontSize: 16, height: 1.5),
            code: TextStyle(
              backgroundColor: Colors.grey.shade200,
              fontFamily: 'monospace',
            ),
          ),
          maxWidth: 500,
          minWidth: 200,
          flex: 8,
          padding: EdgeInsets.all(16),
          margin: EdgeInsets.symmetric(vertical: 8, horizontal: 4),
        ),
      ),
    );
  }
}
```

--------------------------------

### Default LlmChatViewStrings Instance

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmChatViewStrings.md

Provides a static default instance of LlmChatViewStrings with standard English strings. Useful for quick setup.

```dart
static const LlmChatViewStrings defaults = LlmChatViewStrings()
```

```dart
final defaultStrings = LlmChatViewStrings.defaults;
print(defaultStrings.send);  // 'Send'
```

--------------------------------

### Example Usage of Custom LlmChatViewStrings

Source: https://github.com/flutter/ai/blob/main/_autodocs/types.md

Provide custom string values when creating an instance of LlmChatViewStrings to override default labels. This custom instance can then be passed to LlmChatView.

```dart
final customStrings = LlmChatViewStrings(
  addAttachment: 'Ajouter une pièce jointe',
  typeAMessage: 'Tapez votre message...', 
);

LlmChatView(
  provider: provider,
  strings: customStrings,
)
```

--------------------------------

### Message Sender Example

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/ChatViewModel.md

Use a custom messageSender to augment prompts or add logging before sending messages to the LLM. This overrides the default provider's sendMessageStream method.

```dart
messageSender: (prompt, {required attachments}) {
  final augmented = 'You are a helpful assistant.\n\nUser: $prompt';
  return provider.sendMessageStream(augmented, attachments: attachments);
}
```

--------------------------------

### Get Chat History

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmProvider.md

Example of how to retrieve and iterate through the current chat history.

```dart
for (final message in provider.history) {
  print('${message.origin}: ${message.text}');
}
```

--------------------------------

### LlmChatView with Welcome Message and Suggestions

Source: https://github.com/flutter/ai/blob/main/_autodocs/README.md

Configure a welcome message and initial user suggestions for the LlmChatView.

```dart
return Scaffold(
  body: LlmChatView(
    provider: provider,
    welcomeMessage: 'Welcome to AI Assistant!',
    suggestions: ['Hello', 'What can you do?'],
  ),
);
```

--------------------------------

### Add Dependencies to pubspec.yaml

Source: https://github.com/flutter/ai/blob/main/README.md

Install the required packages for Firebase AI integration in your Flutter project.

```sh
$ flutter pub add flutter_ai_toolkit firebase_ai firebase_core
```

--------------------------------

### LlmChatViewStyle Constructor

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmChatViewStyle.md

Demonstrates how to create a fully customized `LlmChatViewStyle` by providing values for various styling properties, including message appearance, input field, and spacing.

```APIDOC
## LlmChatViewStyle Constructor

### Description
Initializes a new `LlmChatViewStyle` with custom visual properties for the chat view.

### Parameters
- **backgroundColor** (Color) - The background color of the chat view.
- **menuColor** (Color) - The color of the menu elements.
- **progressIndicatorColor** (Color) - The color of the progress indicator.
- **userMessageStyle** (UserMessageStyle) - Styling for user messages.
- **llmMessageStyle** (LlmMessageStyle) - Styling for LLM messages.
- **chatInputStyle** (ChatInputStyle) - Styling for the chat input field.
- **messageSpacing** (double) - Spacing between messages.
- **padding** (EdgeInsetsGeometry) - Padding around the chat view content.
- **margin** (EdgeInsetsGeometry) - Margin around the chat view.
- **suggestionStyle** (SuggestionStyle) - Styling for suggestion chips.
- **submitButtonStyle** (ActionButtonStyle) - Styling for the submit button.
- **stopButtonStyle** (ActionButtonStyle) - Styling for the stop button.
- **recordButtonStyle** (ActionButtonStyle) - Styling for the record button.

### Example
```dart
final customStyle = LlmChatViewStyle(
  backgroundColor: Color(0xFFF5F5F5),
  menuColor: Colors.white,
  progressIndicatorColor: Colors.blue,
  userMessageStyle: UserMessageStyle(
    textStyle: TextStyle(fontSize: 16),
    decoration: BoxDecoration(color: Colors.blue.shade100, borderRadius: BorderRadius.circular(12)),
  ),
  llmMessageStyle: LlmMessageStyle(
    textStyle: TextStyle(fontSize: 16),
    decoration: BoxDecoration(color: Colors.grey.shade200, borderRadius: BorderRadius.circular(12)),
  ),
  chatInputStyle: ChatInputStyle(
    hintText: 'Type your message...', 
    backgroundColor: Colors.white,
    decoration: BoxDecoration(border: Border.all(color: Colors.grey), borderRadius: BorderRadius.circular(8)),
  ),
  messageSpacing: 12.0,
  padding: EdgeInsets.all(16),
  margin: EdgeInsets.symmetric(vertical: 8),
  suggestionStyle: SuggestionStyle(
    textStyle: TextStyle(fontSize: 14),
    decoration: BoxDecoration(color: Colors.blue.shade50, borderRadius: BorderRadius.circular(20), border: Border.all(color: Colors.blue)),
  ),
  submitButtonStyle: ActionButtonStyle(icon: Icons.send, iconColor: Colors.blue),
  stopButtonStyle: ActionButtonStyle(icon: Icons.stop_circle, iconColor: Colors.red),
  recordButtonStyle: ActionButtonStyle(icon: Icons.mic, iconColor: Colors.red),
);
```
```

--------------------------------

### Initialize EchoProvider

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/EchoProvider.md

Instantiate the EchoProvider, optionally with initial chat history.

```dart
final provider = EchoProvider();

// Or with initial history
final saved = await loadSavedMessages();
final provider = EchoProvider(history: saved);
```

--------------------------------

### Handle LlmCancelException in a Stream

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmException.md

Example of how to catch and handle a LlmCancelException when processing a message stream. This is useful for user-initiated cancellations.

```dart
import 'package:flutter_ai_toolkit/flutter_ai_toolkit.dart';

try {
  await provider.sendMessageStream('Prompt').toList();
} catch (e) {
  if (e is LlmCancelException) {
    print('User cancelled the operation');
  }
}
```

--------------------------------

### Create Basic LlmChatView

Source: https://github.com/flutter/ai/blob/main/_autodocs/README.md

Instantiate the LlmChatView widget, providing the initialized LLM provider. This sets up the basic chat interface.

```dart
LlmChatView(
  provider: provider,
  suggestions: ['Hello', 'How are you?'],
  welcomeMessage: 'Welcome to AI Chat!',
)
```

--------------------------------

### Handle LlmFailureException in a Stream

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmException.md

Example of catching and handling a LlmFailureException, printing the specific error message. This is useful for diagnosing operational failures.

```dart
try {
  await provider.sendMessageStream('Prompt').toList();
} catch (e) {
  if (e is LlmFailureException) {
    print('Operation failed: ${e.message}');
  }
}
```

--------------------------------

### Accessing and Customizing Default Styles

Source: https://github.com/flutter/ai/blob/main/_autodocs/configuration.md

Demonstrates how to obtain the default style configuration and then override specific properties to create a custom style.

```dart
final defaults = LlmChatViewStyle.defaultStyle();

// Customize specific property
final custom = defaults.copyWith(
  backgroundColor: Colors.black,
  messageSpacing: 16.0,
);
```

--------------------------------

### Instantiate LlmChatView

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/ChatViewModel.md

Pass configuration parameters to LlmChatView for it to create a ChatViewModel.

```dart
LlmChatView(
  provider: myProvider,
  style: myStyle,
  suggestions: ['Hello', 'Help'],
  welcomeMessage: 'Welcome!',
  responseBuilder: customResponseBuilder,
  messageSender: customMessageSender,
  speechToText: customSpeechToText,
  enableAttachments: true,
  enableVoiceNotes: true,
  strings: customStrings,
)
```

--------------------------------

### Get Default Light Theme Style

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmChatViewStyle.md

Returns the default light theme style for the chat view. All properties are initialized with sensible defaults.

```dart
factory LlmChatViewStyle.defaultStyle()
```

--------------------------------

### Default ActionButtonStyle Factory

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/ActionButtonStyle.md

Creates a default styled button for a specified action type. Use this to get pre-defined styles for common actions.

```dart
factory ActionButtonStyle.defaultStyle(
  ActionButtonType type, {
  LlmChatViewStrings? strings,
})
```

--------------------------------

### ChatViewModel Constructor

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/ChatViewModel.md

The constructor for ChatViewModel initializes the chat interface with various configurations including the LLM provider, styling, suggestions, welcome message, response builder, message sender, speech-to-text converter, attachment and voice note enablement, and localized strings.

```APIDOC
## ChatViewModel Constructor

### Description
Initializes the ChatViewModel with all necessary state and configuration for the LlmChatView widget.

### Parameters
All parameters are required positional arguments.

- **provider** (LlmProvider) - Required - The LLM provider for the chat interface.
- **style** (LlmChatViewStyle?) - Required - Optional style configuration for the chat view.
- **suggestions** (List<String>) - Required - Predefined suggestions to display when the chat history is empty.
- **welcomeMessage** (String?) - Required - Optional welcome message displayed when the chat is first opened.
- **responseBuilder** (ResponseBuilder?) - Required - Optional custom builder for rendering LLM responses.
- **messageSender** (LlmStreamGenerator?) - Required - Optional custom message sender.
- **speechToText** (SpeechToTextConverter?) - Required - Optional custom speech-to-text converter for voice input.
- **enableAttachments** (bool) - Required - Whether file and image attachments are enabled.
- **enableVoiceNotes** (bool) - Required - Whether voice recording is enabled.
- **strings** (LlmChatViewStrings) - Optional - Localized strings used throughout the chat interface. Defaults to `const LlmChatViewStrings()`.
```

--------------------------------

### ImageFileAttachment Class Definition

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/Attachment.md

Represents an image file attachment, extending FileAttachment. It enforces that the MIME type must start with 'image/' through an assertion.

```dart
final class ImageFileAttachment extends FileAttachment {
  ImageFileAttachment({
    required String name,
    required String mimeType,
    required Uint8List bytes,
  }) : assert(Attachment._isImage(mimeType));
}
```

--------------------------------

### Creating Chat Messages

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/MessageOrigin.md

Shows how to create user and LLM messages using factory constructors and how to access and check their origins.

```dart
import 'package:flutter_ai_toolkit/flutter_ai_toolkit.dart';

// User message using factory constructor
final userMsg = ChatMessage.user('Hello, AI!', []);

// LLM message using factory constructor
final llmMsg = ChatMessage.llm();

// Access origin
print(userMsg.origin);        // MessageOrigin.user
print(llmMsg.origin);         // MessageOrigin.llm

// Check origin
if (userMsg.origin.isUser) {
  print('This is a user message');
}

if (llmMsg.origin.isLlm) {
  print('This is an LLM message');
}
```

--------------------------------

### Validating Conversation Flow Logic

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/MessageOrigin.md

A function to validate if a conversation history follows the correct alternating pattern of user and LLM messages, starting with the user.

```dart
bool isValidConversation(Iterable<ChatMessage> history) {
  if (history.isEmpty) return true;

  var lastOrigin = MessageOrigin.llm; // User goes first
  
  for (final msg in history) {
    // Messages must alternate between user and LLM
    if (msg.origin == lastOrigin) {
      return false;
    }
    lastOrigin = msg.origin;
  }
  
  return true;
}
```

--------------------------------

### LlmChatView Constructor Parameters

Source: https://github.com/flutter/ai/blob/main/_autodocs/MANIFEST.md

Provides a reference to the constructor parameters for the LlmChatView widget. Refer to LlmChatView.md for a detailed table of over 30 options.

```dart
See [LlmChatView.md](./api-reference/LlmChatView.md) - Parameters table (30+ options)
```

--------------------------------

### UserMessageStyle Constructor

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/UserMessageStyle.md

Initializes a UserMessageStyle object with optional text style and decoration.

```APIDOC
## UserMessageStyle()

### Description
Initializes a UserMessageStyle object with optional text style and decoration.

### Parameters
#### Path Parameters
None

#### Query Parameters
None

#### Request Body
None

### Constructor
```dart
const UserMessageStyle({
  this.textStyle,
  this.decoration,
})
```

### Properties
- **textStyle** (TextStyle?) - Optional - Text style for user message content
- **decoration** (Decoration?) - Optional - Background and border decoration
```

--------------------------------

### Handle LlmFailureException in Dart

Source: https://github.com/flutter/ai/blob/main/_autodocs/errors.md

Catch LlmFailureException to handle errors that are not due to cancellation. This example demonstrates accessing the error message property for more specific error reporting.

```dart
try {
  await provider.sendMessageStream('Prompt').toList();
} on LlmFailureException catch (e) {
  print('Operation failed: ${e.message}');
  // Error message examples:
  // 'Network connection timeout'
  // 'Authentication failed'
  // 'Rate limit exceeded'
  // 'Invalid API key'
}
```

--------------------------------

### Implement Custom LLM Provider in Dart

Source: https://github.com/flutter/ai/blob/main/_autodocs/README.md

Demonstrates how to create a custom LLM provider by extending the LlmProvider interface and implementing necessary methods for streaming responses and managing chat history.

```dart
class MyProvider extends LlmProvider with ChangeNotifier {
  final List<ChatMessage> _history = [];

  @override
  Stream<String> generateStream(
    String prompt, {
    Iterable<Attachment> attachments = const {},
  }) async*
 {
    // Generate response
    yield 'Response text';
  }

  @override
  Stream<String> sendMessageStream(
    String prompt,
    {
    Iterable<Attachment> attachments = const {},
  }
  ) async*
 {
    final userMsg = ChatMessage.user(prompt, attachments);
    final llmMsg = ChatMessage.llm();
    _history.addAll([userMsg, llmMsg]);

    yield* generateStream(prompt, attachments: attachments).map((chunk) {
      llmMsg.append(chunk);
      return chunk;
    });

    notifyListeners();
  }

  @override
  Iterable<ChatMessage> get history => _history;

  @override
  set history(Iterable<ChatMessage> messages) {
    _history.clear();
    _history.addAll(messages);
    notifyListeners();
  }
}

```

--------------------------------

### Enable Network Access on macOS

Source: https://github.com/flutter/ai/blob/main/README.md

Add the 'com.apple.security.network.client' key to your *.entitlements file to enable network access for your macOS application.

```xml
<plist version="1.0">
    <dict>
      ...
      <key>com.apple.security.network.client</key>
      <true/>
    </dict>
</plist>
```

--------------------------------

### Custom LlmProvider with Exception Handling

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmException.md

Demonstrates how to implement a custom LlmProvider and handle various exceptions by throwing appropriate LlmException types. This is useful for managing errors gracefully in custom API integrations.

```dart
class CustomProvider extends LlmProvider {
  @override
  Stream<String> sendMessageStream(
    String prompt, {
    Iterable<Attachment> attachments = const [],
  }) async*
 {
    try {
      // Perform API call
      final response = await apiClient.sendPrompt(prompt);
      
      // Yield response chunks
      for (final chunk in response.chunks) {
        yield chunk;
      }
    } on OperationCanceledException {
      throw const LlmCancelException();
    } on ApiException catch (e) {
      throw LlmFailureException('API error: ${e.message}');
    } on SocketException {
      throw const LlmFailureException('Network connection failed');
    }
  }
}
```

--------------------------------

### Custom Error Handling with LlmChatView

Source: https://github.com/flutter/ai/blob/main/_autodocs/errors.md

Provide a custom onErrorCallback to LlmChatView for specific handling of LlmFailureException. This example shows how to parse the error message and display a tailored alert dialog.

```dart
LlmChatView(
  provider: provider,
  onErrorCallback: (context, error) {
    if (error is LlmFailureException) {
      // Specific error handling
      final message = switch (error.message) {
        String m when m.contains('timeout') => 'Request took too long',
        String m when m.contains('auth') => 'Authentication failed',
        String m when m.contains('rate') => 'Too many requests',
        _ => 'An error occurred: ${error.message}',
      };
      
      showDialog(
        context: context,
        builder: (context) => AlertDialog(
          title: Text('Error'),
          content: Text(message),
          actions: [
            TextButton(
              onPressed: () => Navigator.pop(context),
              child: Text('Retry'),
            ),
          ],
        ),
      );
    }
  },
)
```

--------------------------------

### LlmChatView with Default Strings

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmChatViewStrings.md

Demonstrates how to use the LlmChatView widget without providing custom strings, which will automatically use the default set.

```dart
LlmChatView(
  provider: myProvider,
  // Uses LlmChatViewStrings.defaults if not provided
)
```

--------------------------------

### Handle LlmCancelException in Dart

Source: https://github.com/flutter/ai/blob/main/_autodocs/errors.md

Use a try-catch block to specifically catch and handle LlmCancelException, which is thrown when an LLM operation is cancelled. This example shows how to print a message when cancellation occurs.

```dart
try {
  await provider.sendMessageStream('Prompt').toList();
} on LlmCancelException {
  print('User cancelled the operation');
}
```

--------------------------------

### ActionButtonStyle Constructor

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/ActionButtonStyle.md

Initializes an ActionButtonStyle with optional icon, color, decoration, text, and text style properties.

```APIDOC
## ActionButtonStyle Constructor

```dart
const ActionButtonStyle({
  this.icon,
  this.iconColor,
  this.iconDecoration,
  this.text,
  this.textStyle,
})
```

### Properties

| Property | Type | Description |
|----------|------|-------------|
| icon | IconData? | Button icon to display |
| iconColor | Color? | Color of the icon |
| iconDecoration | Decoration? | Background decoration around icon |
| text | String? | Tooltip or label text |
| textStyle | TextStyle? | Text appearance |
```

--------------------------------

### Widget Test for LlmChatView Welcome Message

Source: https://github.com/flutter/ai/blob/main/_autodocs/README.md

Write a widget test to verify that the LlmChatView displays the correct welcome message when using EchoProvider.

```dart
testWidgets('Chat displays welcome message', (tester) async {
  await tester.pumpWidget(
    MaterialApp(
      home: Scaffold(
        body: LlmChatView(
          provider: EchoProvider(),
          welcomeMessage: 'Welcome!',
        ),
      ),
    ),
  );

  expect(find.text('Welcome!'), findsOneWidget);
});
```

--------------------------------

### Replace Chat History

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/FirebaseProvider.md

Replaces the entire chat history with a new set of messages, effectively starting a new conversation session. This setter also notifies listeners of the change.

```dart
set history(Iterable<ChatMessage> history)
```

```dart
// Restore a saved conversation
final savedMessages = await loadConversation('conversation-id');
provider.history = savedMessages;

// Clear the conversation
provider.history = [];
```

--------------------------------

### Custom User Message Styling in LlmChatView

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/UserMessageStyle.md

Example of how to apply a custom UserMessageStyle within an LlmChatView. This demonstrates overriding text style and decoration properties like color and border radius.

```dart
import 'package:flutter/material.dart';
import 'package:flutter_ai_toolkit/flutter_ai_toolkit.dart';

class CustomChatView extends StatelessWidget {
  @override
  Widget build(BuildContext context) {
    return LlmChatView(
      provider: myProvider,
      style: LlmChatViewStyle(
        userMessageStyle: UserMessageStyle(
          textStyle: TextStyle(
            fontSize: 16,
            color: Colors.white,
            fontWeight: FontWeight.w500,
          ),
          decoration: BoxDecoration(
            color: Colors.blue,
            borderRadius: BorderRadius.only(
              topLeft: Radius.circular(20),
              topRight: Radius.circular(4),
              bottomLeft: Radius.circular(20),
              bottomRight: Radius.circular(20),
            ),
            boxShadow: [
              BoxShadow(
                color: Colors.blue.withOpacity(0.3),
                blurRadius: 8,
                offset: Offset(0, 2),
              ),
            ],
          ),
        ),
      ),
    );
  }
}
```

--------------------------------

### EchoProvider Constructor

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/EchoProvider.md

Initializes a new instance of the EchoProvider. It can optionally be initialized with a pre-existing chat history.

```APIDOC
## EchoProvider Constructor

### Description
Initializes a new instance of the EchoProvider. It can optionally be initialized with a pre-existing chat history.

### Parameters
#### Path Parameters
None

#### Query Parameters
None

#### Request Body
None

### Constructor Signature
```dart
EchoProvider({Iterable<ChatMessage>? history})
```

### Parameters
- **history** (Iterable<ChatMessage>) - Optional - Optional initial chat history

### Example
```dart
final provider = EchoProvider();

// Or with initial history
final saved = await loadSavedMessages();
final provider = EchoProvider(history: saved);
```
```

--------------------------------

### Directory Structure Overview

Source: https://github.com/flutter/ai/blob/main/_autodocs/MANIFEST.md

Provides a visual representation of the documentation's directory structure, showing the location of various reference files.

```markdown
output/
├── README.md                 # Start here - overview and quick links
├── MANIFEST.md              # This file - navigation and structure
├── types.md                 # Type definitions
├── configuration.md         # Configuration options
├── errors.md                # Error handling
└── api-reference/           # Individual component references
    ├── LlmChatView.md
    ├── LlmProvider.md
    ├── ChatMessage.md
    ├── ChatViewModel.md
    ├── FirebaseProvider.md
    ├── EchoProvider.md
    ├── Attachment.md
    ├── MessageOrigin.md
    ├── LlmException.md
    ├── LlmChatViewStyle.md
    ├── UserMessageStyle.md
    ├── LlmMessageStyle.md
    ├── ActionButtonStyle.md
    └── LlmChatViewStrings.md
```

--------------------------------

### Initialize FirebaseProvider

Source: https://github.com/flutter/ai/blob/main/_autodocs/README.md

Initialize the FirebaseProvider with a specific generative model. This is the first step to integrating AI chat functionality.

```dart
final provider = FirebaseProvider(
  model: FirebaseAI.vertexAI().generativeModel(
    model: 'gemini-2.0-flash',
  ),
);
```

--------------------------------

### Implementing a Logging Error Wrapper Provider

Source: https://github.com/flutter/ai/blob/main/_autodocs/errors.md

Create a wrapper provider like LoggingProvider to intercept and log errors from an inner LlmProvider before re-throwing them. This helps in debugging and monitoring without altering the core error handling logic.

```dart
class LoggingProvider extends LlmProvider {
  final LlmProvider _inner;

  @override
  Stream<String> sendMessageStream(
    String prompt,
    {
    Iterable<Attachment> attachments = const [],
  }
  ) async*
 {
    try {
      yield* _inner.sendMessageStream(prompt, attachments: attachments);
    } catch (e) {
      // Log the error
      logger.error('LLM error: $e', stackTrace: StackTrace.current);
      // Re-throw for caller to handle
      rethrow;
    }
  }
}
```

--------------------------------

### LlmMessageStyle Constructor

Source: https://github.com/flutter/ai/blob/main/_autodocs/api-reference/LlmMessageStyle.md

Configures the visual style for LLM messages, including icons, background decorations, text markdown styling, and layout constraints.

```APIDOC
## LlmMessageStyle()

### Description

Constructs a style configuration for Language Model messages, allowing customization of visual elements and layout behavior.

### Parameters

#### Named Parameters

- **icon** (IconData?): Optional. The leading icon for LLM messages. Defaults to null.
- **iconColor** (Color?): Optional. The color of the icon. Defaults to null.
- **iconDecoration** (Decoration?): Optional. The background decoration applied around the icon. Defaults to null.
- **decoration** (Decoration?): Optional. The background and border decoration for the message bubble. Defaults to null.
- **markdownStyle** (MarkdownStyleData?): Optional. Configuration for rendering Markdown within the message. Defaults to null.
- **maxWidth** (double): Optional. The maximum width constraint for the message bubble. Defaults to 600.0.
- **minWidth** (double): Optional. The minimum width constraint for the message bubble. Defaults to 300.0.
- **flex** (int): Optional. The flex factor used in a row layout to determine relative sizing. Defaults to 6.
- **padding** (EdgeInsetsGeometry): Optional. The inner padding within the message bubble. Defaults to `EdgeInsets.all(12.0)`.
- **margin** (EdgeInsetsGeometry): Optional. The outer margin around the message bubble. Defaults to `EdgeInsets.symmetric(vertical: 4.0)`.
```