mcpField
S
syncfusion-flutter-chat

syncfusion-flutter-chat

github.com/syncfusionDevelopment255 installsUpdated 17w ago
Repository

About syncfusion-flutter-chat

A reference skill for building conversational interfaces in Flutter with Syncfusion's SfChat and SfAIAssistView. It helps developers choose between multi-user messaging and AI request-response UI, and covers message composition, customization, and RTL support.

Key features

  1. 01Multi-user chat interfaces
  2. 02AI request-response interfaces
  3. 03Custom messages and composers
  4. 04Suggestions and response actions
  5. 05Themes and empty-state placeholders
  6. 06Right-to-left layout support

Use cases

  1. 01Building team or customer-support chat
  2. 02Creating an AI chatbot interface
  3. 03Adding response loading and toolbar actions
  4. 04Localizing conversation UI for RTL

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/syncfusion/flutter-ui-components-skills --skill syncfusion-flutter-chat

After installing, type/syncfusion-flutter-chatin Claude Code to confirm it loaded.

Bundled files

References and scripts installed along with SKILL.md.

references/
action-button.md
aiassistview-overview.md
chat-overview.md
common-patterns.md
composer.md
conversation-area.md
getting-started.md
placeholder.md
right-to-left.md
theming.md

Details

From the project README

Syncfusion Flutter Chat & AI AssistView

This skill covers two related Syncfusion Flutter components for conversational UI: SfChat (Chat widget) and SfAIAssistView (AI AssistView widget). While they share many visual and configuration features, they serve different primary purposes.

When to Use This Skill

Use this skill when you need to:

  • Implement chat interfaces with multi-user conversations and messaging
  • Add AI assistant functionality with request-response patterns
  • Build messaging apps with incoming/outgoing message support
  • Create AI chatbot interfaces with AI service integration
  • Display conversation UI with messages, avatars, and timestamps
  • Handle message composition with text editors and action buttons
  • Customize message appearance with builders, themes, and styling
  • Support placeholder screens for empty conversation states
  • Enable suggestion chips for quick responses or actions
  • Integrate toolbar actions for response messages (like, copy, retry)
  • Show loading indicators for AI response generation
  • Implement RTL support for right-to-left languages

Choosing the Right Component

Use SfChat (Chat Widget) when:
  • You need multi-user conversation interfaces
  • Your app requires incoming and outgoing messages differentiation
  • You need traditional messaging app functionality (WhatsApp, Telegram style)
  • Building chat for: team collaboration, customer support, social messaging, group chats
  • Message suggestions should appear on both incoming/outgoing messages
  • You need to identify the outgoing user explicitly
Use SfAIAssistView (AI AssistView Widget) when:
  • You need AI assistant or chatbot functionality
  • Your primary goal is request-response patterns with AI services
  • You need toolbar items on response messages (like, dislike, copy, retry)
  • You need loading indicators while AI generates responses
  • Building UI for: AI assistants, chatbots, help bots, Q&A interfaces
  • Message suggestions should only appear on AI response messages
  • You need clear request vs response message distinction
Key Differences Summary:
FeatureSfChatSfAIAssistView
Primary PurposeMulti-user messagingAI assistant interaction
Message TypesIncoming/OutgoingRequest/Response
User IdentificationoutgoingUser propertyAuthor in each message
SuggestionsOn any messageOnly on response messages
Toolbar Items❌ Not supported✅ Response toolbar support
Loading Indicator❌ Not included✅ Response loading builder
Use CaseChat apps, messagingAI chatbots, assistants
Message ClassesChatMessageAssistMessage
Composer ClassesChatComposerAssistComposer
Action Button ClassesChatActionButtonAssistActionButton

Documentation and Navigation Guide

Getting Started

📄 Read: references/getting-started.md

  • Installation and package setup for both components
  • Basic SfChat implementation
  • Basic SfAIAssistView implementation
  • Package dependencies and imports
  • Quick comparison and first examples
Component Overview

📄 For Chat: references/chat-overview.md

  • SfChat widget overview and features
  • When to use Chat vs AI AssistView
  • Multi-user conversation capabilities
  • Incoming/outgoing message patterns
  • Chat-specific capabilities
  • Basic chat configuration

📄 For AI AssistView: references/aiassistview-overview.md

  • SfAIAssistView widget overview and features
  • When to use AI AssistView vs Chat
  • Request/response patterns
  • AI integration capabilities
  • AIAssistView-specific capabilities
  • Basic AIAssistView configuration
Conversation Area and Messages

📄 Read: references/conversation-area.md

  • Message structure and display
  • Incoming/outgoing messages (Chat)
  • Request/response messages (AIAssistView)
  • Message settings and customization
  • Headers and timestamps
  • Footers and additional info
  • Message avatars
  • Content area styling
  • Message suggestions (both components)
  • Toolbar items (AIAssistView only)
  • Loading indicators (AIAssistView only)
Composer and Input

📄 Read: references/composer.md

  • Default composer configuration
  • Text editor customization
  • Minimum and maximum lines
  • Decoration and styling
  • Hint text and placeholders
  • Borders and padding
  • Prefix and suffix icons
  • Text style customization
  • Margin configuration
  • Custom composer builder
Action Button

📄 Read: references/action-button.md

  • Action button overview
  • Send button functionality
  • Custom child widgets
  • onPressed callback
  • Tooltip text
  • Colors (foreground, background, focus, hover, splash)
  • Elevation settings
  • Mouse cursor customization
  • Shape and border radius
  • Margin and size
Placeholder

📄 Read: references/placeholder.md

  • Placeholder builder overview
  • Custom placeholder widgets
  • Empty conversation state design
  • Welcome messages
  • Getting started hints
  • Examples for both components
Theming and Customization

📄 Read: references/theming.md

  • Theme overview (Chat vs AIAssistView)
  • SfChatTheme and SfAIAssistViewTheme
  • Action button theming
  • Avatar colors and styling
  • Message background colors
  • Content text styles
  • Header text styles
  • Suggestion styling
  • Message shapes and borders
  • Toolbar theming (AIAssistView only)
Localization and Accessibility

📄 Read: references/right-to-left.md

  • Right-to-left (RTL) support
  • Directionality widget usage
  • RTL rendering for all elements
  • RTL for placeholder
  • RTL for composer
  • RTL for action button
  • RTL for messages and content
Common Patterns

📄 Read: references/common-patterns.md

  • Message settings customization
  • Custom composer with multiple actions
  • Custom placeholder for empty state
  • Loading indicator for AI responses
  • Theming for consistent brand experience

Quick Start Examples

Basic Chat with Messages
import 'package:flutter/material.dart';
import 'package:syncfusion_flutter_chat/chat.dart';

class MyChat extends StatefulWidget {
  @override
  State<MyChat> createState() => _MyChatState();
}

class _MyChatState extends State<MyChat> {
  List<ChatMessage> _messages = <ChatMessage>[];

  @override
  void initState() {
    super.initState();
    _messages = <ChatMessage>[
      ChatMessage(
        text: 'Hi! How can I help you today?',
        time: DateTime.now(),
        author: const ChatAuthor(
          id: '123-001',
          name: 'John Doe',
        ),
      ),
      ChatMessage(
        text: 'I need help with my order.',
        time: DateTime.now(),
        author: const ChatAuthor(
          id: '123-002',
          name: 'Jane Smith',
        ),
      ),
    ];
  }

  @override
  Widget build(BuildContext context) {
    return Scaffold(
      appBar: AppBar(title: Text('Chat')),
      body: SfChat(
        messages: _messages,
        outgoingUser: '123-002',
        composer: const ChatComposer(
          decoration: InputDecoration(
            hintText: 'Type a message',
          ),
        ),
        actionButton: ChatActionButton(
          onPressed: (String newMessage) {
            setState(() {
              _messages.add(
                ChatMessage(
                  text: newMessage,
                  time: DateTime.now(),
                  author: const ChatAuthor(
                    id: '123-002',
                    name: 'Jane Smith',
                  ),
                ),
              );
            });
          },
        ),
      ),
    );
  }
}
Basic AI AssistView with AI Integration
import 'package:flutter/material.dart';
import 'package:syncfusion_flutter_chat/assist_view.dart';

class MyAIAssistView extends StatefulWidget {
  @override
  State<MyAIAssistView> createState() => _MyAIAssistViewState();
}

class _MyAIAssistViewState extends State<MyAIAssistView> {
  List<AssistMessage> _messages = <AssistMessage>[];

  void _generateAIResponse(String userRequest) async {
    // Call your AI service here
    final String aiResponse = await _getAIResponse(userRequest);
    setState(() {
      _messages.add(
        AssistMessage.response(
          data: aiResponse,
          time: DateTime.now(),
          author: const AssistMessageAuthor(
            id: 'ai-001',
            name: 'AI Assistant',
          ),
        ),
      );
    });
  }

  Future<String> _getAIResponse(String request) async {
    // Connect with your preferred AI service
    // Example: OpenAI, Google AI, Azure AI, etc.
    await Future.delayed(Duration(seconds: 2)); // Simulate API call
    return 'This is a response to: $request';
  }

  @override
  Widget build(BuildContext context) {
    return Scaffold(
      appBar: AppBar(title: Text('AI Assistant')),
      body: SfAIAssistView(
        messages: _messages,
        composer: const AssistComposer(
          decoration: InputDecoration(
            hintText: 'Ask me anything',
          ),
        ),
        actionButton: AssistActionButton(
          onPressed: (String userRequest) {
            setState(() {
              _messages.add(
                AssistMessage.request(
                  data: userRequest,
                  time: DateTime.now(),
                  author: const AssistMessageAuthor(
                    id: 'user-001',
                    name: 'User',
                  ),
                ),
              );
              _generateAIResponse(userRequest);
            });
          },
        ),
        placeholderBuilder: (BuildContext context) {
          return const Center(
            child: Text(
              'What can I help you with today?',
              style: TextStyle(fontSize: 16),
            ),
          );
        },
      ),
    );
  }
}
Chat with Suggestions
SfChat(
  messages: <ChatMessage>[
    ChatMessage(
      text: 'Which programming language should I learn?',
      time: DateTime.now(),
      author: const ChatAuthor(id: '1', name: 'Alice'),
    ),
    ChatMessage(
      text: 'It depends on your goals. What interests you?',
      time: DateTime.now(),
      author: const ChatAuthor(id: '2', name: 'Bob'),
      suggestions: <ChatMessageSuggestion>[
        const ChatMessageSuggestion(data: 'Web Development'),
        const ChatMessageSuggestion(data: 'Mobile Apps'),
        const ChatMessageSuggestion(data: 'Data Science'),
        const ChatMessageSuggestion(data: 'Game Development'),
      ],
    ),
  ],
  outgoingUser: '1',
)
AI AssistView with Toolbar Items
void _generateResponse(String request) async {
  final String response = await _getAIResponse(request);
  setState(() {
    _messages.add(
      AssistMessage.response(
        data: response,
        time: DateTime.now(),
        author: const AssistMessageAuthor(id: 'ai', name: 'AI'),
        toolbarItems: <AssistMessageToolbarItem>[
          const AssistMessageToolbarItem(
            content: Icon(Icons.thumb_up_outlined),
            tooltip: 'Like',
          ),
          const AssistMessageToolbarItem(
            content: Icon(Icons.thumb_down_outlined),
            tooltip: 'Dislike',
          ),
          const AssistMessageToolbarItem(
            content: Icon(Icons.copy),
            tooltip: 'Copy',
          ),
          const AssistMessageToolbarItem(
            content: Icon(Icons.restart_alt),
            tooltip: 'Regenerate',
          ),
        ],
      ),
    );
  });
}

SfAIAssistView(
  messages: _messages,
  onToolbarItemSelected: (bool selected, int messageIndex,
      AssistMessageToolbarItem item, int toolbarItemIndex) {
    // Handle toolbar action (like, dislike, copy, etc.)
    print('Toolbar item selected at index $toolbarItemIndex');
  },
)

Key Properties

SfChat Essential Properties
  • messages - List of ChatMessage objects to display
  • outgoingUser - ID of the user sending messages (distinguishes incoming/outgoing)
  • composer - ChatComposer for text input
  • actionButton - ChatActionButton for sending messages
  • placeholderBuilder - Custom widget for empty state
  • incomingMessageSettings - Settings for incoming messages
  • outgoingMessageSettings - Settings for outgoing messages
  • messageHeaderBuilder - Custom header for each message
  • messageFooterBuilder - Custom footer for each message
  • messageContentBuilder - Custom content for each message
  • messageAvatarBuilder - Custom avatar for each message
SfAIAssistView Essential Properties
  • messages - List of AssistMessage objects (request/response)
  • composer - AssistComposer for text input
  • actionButton - AssistActionButton for sending requests
  • placeholderBuilder - Custom widget for empty state
  • requestMessageSettings - Settings for request messages
  • responseMessageSettings - Settings for response messages
  • messageHeaderBuilder - Custom header for each message
  • messageFooterBuilder - Custom footer for each message
  • messageContentBuilder - Custom content for each message
  • messageAvatarBuilder - Custom avatar for each message
  • responseLoadingBuilder - Custom loading indicator while AI responds
  • onToolbarItemSelected - Callback when toolbar item is tapped
  • onSuggestionItemSelected - Callback when suggestion chip is tapped
ChatMessage / AssistMessage Properties

ChatMessage:

  • text - Message content
  • time - Timestamp of the message
  • author - ChatAuthor with id, name, and avatar
  • suggestions - List of ChatMessageSuggestion items

AssistMessage:

  • data - Message content
  • time - Timestamp of the message
  • author - AssistMessageAuthor with id, name, and avatar
  • suggestions - List of AssistMessageSuggestion items (response only)
  • toolbarItems - List of AssistMessageToolbarItem (response only)
Composer Properties (Both Components)
  • minLines - Minimum lines in text editor (default: 1)
  • maxLines - Maximum lines in text editor (default: 6)
  • decoration - InputDecoration for styling
  • margin - Space around the composer
  • textStyle - Text style for input
  • builder - Custom composer widget
Action Button Properties (Both Components)
  • child - Custom widget for button
  • onPressed - Callback when button is pressed
  • tooltip - Tooltip text
  • foregroundColor - Icon/text color
  • backgroundColor - Button background color
  • focusColor - Color when focused
  • hoverColor - Color when hovered
  • splashColor - Ripple effect color
  • elevation - Shadow elevation
  • focusElevation - Elevation when focused
  • hoverElevation - Elevation when hovered
  • highlightElevation - Elevation when pressed
  • mouseCursor - Cursor style on hover
  • shape - Button shape and border
  • margin - Space around button
  • size - Button dimensions
Message Settings Properties (Both Components)
  • showAuthorName - Show/hide author name
  • showTimestamp - Show/hide timestamp
  • showAuthorAvatar - Show/hide avatar
  • timestampFormat - DateFormat for timestamp
  • backgroundColor - Message background color
  • textStyle - Message text style
  • headerTextStyle - Header text style
  • shape - Message bubble shape
  • widthFactor - Message width relative to screen (0-1)
  • avatarSize - Avatar dimensions
  • margin - Space around message
  • padding - Padding inside message
  • avatarPadding - Padding around avatar
  • headerPadding - Padding around header
  • footerPadding - Padding around footer

Common Use Cases

For SfChat: Customer support chat, team messaging, social messaging, live chat support with multi-user conversations and incoming/outgoing message patterns.

For SfAIAssistView: AI chatbots, AI assistants, help bots, Q&A interfaces with request/response patterns, toolbar items, and loading indicators.

Comments

Markdown supported