Content
# 🌸 Musea/ˈmjuː-zi-ə/ - Intelligent Focus Time Management Assistant
<div align="center">






*A powerful Chrome extension that combines AI virtual characters and the MCP (Model Context Protocol) architecture to help you stay focused and improve work efficiency.*
---
**🌍 Language / 语言**
[](./README_EN.md) | [](./README.md)
---
## 🎯 Target Audience
- **📚 Students** - Need a focused environment for exam preparation, thesis writing, and online learning.
- **💼 Professionals** - Avoid online distractions during remote work, project development, and document writing.
- **🎨 Creators** - Require deep focus for creative work such as writing, design, and programming.
- **📖 Lifelong Learners** - Focus management during skill enhancement, knowledge acquisition, and personal growth.
- **🧘 Focus Trainers** - Anyone looking to cultivate focus habits and improve work efficiency.
## 🧠 The Right Concept of Focus Training
### 💡 Understanding Addiction Mechanisms
Modern social media and entertainment platforms hijack our dopamine systems through **instant feedback** and **random reward** mechanisms, creating a strong psychological dependency. This dependency causes our brains to instinctively seek these "pleasurable stimuli" when we need to focus.
### 🎯 Why We Need to Detox
- **Fragmented Attention** - Frequent information switching disrupts deep thinking abilities.
- **Instant Gratification Trap** - Reduces patience and perseverance towards long-term goals.
- **Cognitive Overload** - The brain is fatigued from processing irrelevant information, affecting work efficiency.
### 🛠️ The Necessity of Tool Assistance
**Pure willpower detox often fails** because:
- Willpower is a limited resource that can easily be depleted.
- Environmental temptations are everywhere, making resistance costly.
- Lack of external structural support makes it easy to falter.
**The Role of Tools**: Provide **external constraints** and **buffer periods**, allowing the brain to gradually adapt to new behavior patterns.
### ⚠️ Avoid Permanent Dependency
**Risks of Tool Dependency**:
- Once the tool is lost, focus immediately regresses.
- Creates "learned helplessness," leading to the belief that one cannot self-regulate.
- When tools fail, rebounds can be even more intense.
### 🌱 Cultivating a Positive Cycle
#### Phase One: Tool Assistance Period (1-3 months)
- Use tools like Musea to establish basic focus habits.
- Gradually extend focus time from 25 minutes to 90 minutes.
- Record focus achievements to build a sense of accomplishment.
#### Phase Two: Active Management Period (3-6 months)
- **Time Planning**: Create detailed schedules to fill time gaps with plans.
- **Interest Substitution**: Cultivate healthy hobbies (exercise, reading, creation).
- **Environment Optimization**: Remove distractions from the physical environment to create a focused atmosphere.
#### Phase Three: Autonomous Focus Period (6 months+)
- **Intrinsic Motivation**: Deeply bind focus with personal values and long-term goals.
- **Positive Feedback**: Gain intrinsic satisfaction from completing meaningful tasks.
- **Habit Solidification**: Make focus a natural behavior pattern.
### 🎯 Core Principles
> **Tools are crutches, not wheelchairs.**
> Use them to help you relearn how to "walk," not to sit forever.
**Remember**: True focus comes from an internal sense of **purpose**, **achievement**, and **self-efficacy**; tools are merely bridges to help you rebuild these capabilities.
</div>
## 📖 Project Overview
Musea is a modern Chrome browser extension designed to enhance work and learning efficiency. It combines the traditional Pomodoro Technique with modern AI technology, providing users with a personalized focus experience and intelligent work assistant through smart virtual characters and the MCP (Model Context Protocol) architecture.
### 🎯 Core Features
#### 🕐 Focus Time Management
- **Smart Focus Timer** - Customizable focus time management (1-120 minutes).
- **Real-time Badge Display** - Dynamically shows remaining time, with color changes over time.
- **Automatic Status Management** - Fully automated start, pause, and stop of focus mode.
#### 🚫 Website Blocking System
- **Complete Blocking Mode** - Completely prevents access to specified websites.
- **Learning Mode** - Hides distracting elements while retaining core learning content.
- **Site-Specific Handling** - Pre-set support for popular websites like Baidu, Bilibili, Zhihu, etc.
#### 🔊 Diverse Notification System
- **Sound Alerts** - Built-in notification sound effects with adjustable volume.
- **TTS Voice Synthesis** - Integrated ByteDance TTS API with various voice types.
- **AI Personalized Notifications** - Warm encouragement reminders generated based on LLM.
#### 🤖 AI Virtual Character Assistant
- **Interactive Virtual Characters** - Cute AI assistants that support click-to-chat.
- **Voice Conversation Feature** - Supports voice input and TTS voice replies.
- **Animation System** - Rich character animations (idle, thinking, greeting, etc.).
- **Conversation Persistence** - Saves conversation history across sessions.
#### 🔗 MCP Intelligent Work Assistant
- **Three-Layer Architecture Design** - Chrome Extension ↔ Native Messaging Host ↔ MCP Client.
- **Academic Research Assistant** - Integrated arxiv paper search functionality.
- **System Control** - Music playback, application launch, system information retrieval.
- **Chrome Search** - Intelligent search and webpage operations.
- **Modular Extensions** - Supports adding more MCP services.
#### 🎨 Modern Interface
- **Responsive Design** - Adapts to different screen sizes.
- **Dark Mode Support** - Eye-friendly dark theme.
- **Animation Effects** - Smooth transitions and interactive feedback.
- **Highly Customizable** - Rich personalization settings options.
## 🚀 Quick Start
### Installation Methods
#### Method 1: Developer Mode Installation (Recommended) ✅
1. **Enable Developer Mode**
- Open the [Chrome Extensions Management Page](chrome://extensions/).
- Turn on "Developer mode" in the upper right corner.
2. **Load the Extension**
- Click "Load unpacked" and select the project's `dist` folder.
3. **Pin the Extension Icon**
- Click the extension icon in the Chrome toolbar.
- Find "Musea" and click to pin it.
4. **Start Using**
- Click the extension icon to open the settings panel.
- Configure your focus time and preferences.
<details>
<summary>📸 Installation Steps Screenshots</summary>
**Load Extension**
<img width="1464" alt="Load Extension" src="https://private-user-images.githubusercontent.com/160222653/438998601-2be2c314-4390-47ed-b5d5-ccdff7c60900.png" />
**Enable Extension**
<img width="426" alt="Enable Extension" src="https://github.com/user-attachments/assets/4c72ae73-dc2d-4a8b-bc14-f80cea76108f" />
**Pin Extension Icon**
<img width="412" alt="Pin Extension" src="https://github.com/user-attachments/assets/f6dc817d-295a-4154-8189-c0b957e08136" />
**Settings Panel**
<img width="458" alt="Settings Panel" src="https://github.com/user-attachments/assets/e46fdc20-507f-4fb9-a369-0bb55eb260d7" />
**Focus Mode Page**
<img width="1467" alt="Focus Mode" src="https://github.com/user-attachments/assets/0c599ec3-11cc-4811-b13b-e0a4cb6e0ea7" />
</details>
#### Method 2: Chrome Web Store ⌛️
> **Note**: For security reasons, there are currently no plans to release on the Chrome Web Store. We adhere to the principles of open-source transparency and encourage users to install directly from the source code!
### Development Environment Setup
#### System Requirements
- **Node.js**: >= 22.12.0
- **Package Manager**: pnpm 9.15.1
- **Browser**: Chrome 109+ or Firefox 109+
#### Build Steps
```bash
# Clone the project
git clone <repository-url>
cd Musea
# Install dependencies
pnpm install
# Development mode build
pnpm dev
# Production mode build
pnpm build
# Package the extension
pnpm zip
```
## 📋 Feature Details
### 🕐 Focus Time Management
- **Flexible Time Settings**: Supports customizable focus durations from 1 to 120 minutes.
- **Smart Badge Display**: Real-time display of remaining focus time, with color changes over time.
- **Automatic Status Management**: Fully automated start, pause, and stop of focus mode.
### 🚫 Website Blocking System
#### Complete Blocking Mode
- Completely prevents access to specified websites.
- Displays a friendly reminder page.
- Supports domain and full URL matching.
#### Learning Mode
- Hides distracting elements on websites (e.g., recommendations, comments).
- Retains core learning content.
- Pre-set support for popular websites like Baidu, Bilibili, etc.
- Custom CSS selector support (use case: for example, when taking screenshots, you may need to hide control bars, buttons, etc., especially when capturing videos).
### 🔊 Diverse Notification System
#### Sound Notifications
- Built-in notification sound effects.
- Adjustable volume (0-100%).
- Test playback feature.
#### TTS Voice Synthesis
- Integrated ByteDance TTS API.
- Multiple voice types available (male/female).
- Adjustable speech rate (0.5x - 2.0x).
- Intelligent caching mechanism to save API calls.
#### AI Personalized Notifications
- Personalized break reminders generated based on LLM.
- Customizable system prompts and user templates.
- Intelligent pre-generation to reduce waiting time.
- Warm and encouraging language style.
### 🤖 AI Virtual Character System
#### Character Interaction Features
- **Smart Dialogue** - Natural language conversation based on LLM.
- **Voice Interaction** - Supports voice input recording and TTS voice replies.
- **Animation Performance** - Rich character animation system (idle, thinking, greeting, speaking, etc.).
- **Context Awareness** - Adjusts behavior based on focus status and website environment.
#### MCP Integration Features
- **Research Assistant** - Calls arxiv paper search through the MCP protocol.
- **System Control** - Music playback, application launch, system information queries.
- **Intelligent Search** - Chrome search and webpage operations.
- **Task Execution** - Automated workflows and task management.
#### Custom Prompt Features
- **System Prompts** - Define the character and behavior norms of the AI assistant.
- **User Templates** - Specific instructions for generating notifications.
- **Placeholder Support** - Supports dynamic variables like `{duration}`.
- **Intelligent Fallback** - Automatically uses default templates when not configured.
#### Default AI Settings
- Friendly and positive language style.
- Concise and effective reminder content (≤50 words).
- Neighborly or cute younger sister tone.
- Includes break suggestions and encouraging phrases.
### 🎨 User Interface
#### Modern Design
- **Responsive Layout**: Adapts to different screen sizes.
- **Dark Mode**: Eye-friendly dark theme.
- **Animation Effects**: Smooth transitions and interactive feedback.
- **Intuitive Operation**: Clear icons and labels.
#### Settings Panel
- **Collapsible Design**: Saves space and expands as needed.
- **Real-time Preview**: Changes take effect immediately.
- **Import/Export**: Configuration backup and restoration (planned).
## 🛠️ Technical Architecture
### Frontend Technology Stack
- **React 19.0.0**: Modern user interface framework.
- **TypeScript 5.8.1**: Type-safe development experience.
- **Tailwind CSS**: Utility-first CSS framework.
- **Vite 6.1.0**: Fast build tool.
### Chrome Extension Architecture
- **Manifest V3**: The latest Chrome extension standard.
- **Service Worker**: Manages background scripts and focus time control.
- **Content Scripts**: Operates page content, website blocking, and virtual character injection.
- **Offscreen Documents**: Supports audio playback and TTS voice synthesis.
- **Native Messaging**: Communicates with local MCP services.
### MCP Three-Layer Architecture
```
Chrome Extension (Frontend) ←→ Native Messaging Host ←→ MCP Client Service ←→ MCP Servers
│ │ │ │
UI Interaction Interface Protocol Bridge MCP Client Implementation Specific Service Implementation
```
- **Chrome Extension Layer**: User interface, virtual character interaction, initiates MCP requests.
- **Native Messaging Host**: Protocol bridge for secure local system access.
- **MCP Client Service**: Service scheduler managing tools and external MCP services.
- **MCP Servers**: Specific functionality implementations (arxiv search, system control, etc.).
### Virtual Character System Architecture
- **Character Service**: Character state management and animation control.
- **Chat System**: Conversation history management and session persistence.
- **Voice Service**: Voice recording, TTS synthesis, audio playback.
- **MCP Integration**: Intelligent task detection and MCP service calls.
### Storage System
- **Chrome Storage API**: Configuration data persistence.
- **Real-time Sync**: State synchronization across tabs.
- **Type Safety**: TypeScript interface definitions.
- **Layered Storage**: Separate management of focus configurations, character settings, and conversation history.
### Package Management
- **Monorepo Architecture**: Modular code organization.
- **Turbo**: Efficient build system.
- **pnpm**: Fast package manager.
## 📁 Project Structure
```
Fuck_Brain_Concentration/
├── chrome-extension/ # Core of the Chrome extension
│ ├── src/
│ │ ├── background/ # Background scripts (focus management, MCP communication)
│ │ ├── content/ # Content scripts (website blocking)
│ │ ├── services/ # Service layer (TTS, character, MCP)
│ │ └── constants/ # Constant definitions
│ ├── public/ # Static resources (audio, icons, pages)
│ └── manifest.ts # Extension manifest
├── MCP/ # MCP backend services
│ ├── native-host/ # Native Messaging Host
│ ├── client/ # MCP client service
│ ├── scripts/ # Installation and startup scripts
│ └── logs/ # Service logs
├── packages/ # Shared packages
│ ├── storage/ # Storage management
│ ├── ui/ # UI components
│ ├── shared/ # Shared utilities
│ └── ...
├── pages/ # Page components
│ ├── popup/ # Popup page (settings panel)
│ ├── content/ # Content page (blocking page)
│ ├── content-ui/ # Content UI (notification cards)
│ └── content-runtime/ # Runtime content (virtual characters)
├── dist/ # Build output
└── local_history/ # Development documentation and history
```
## 🔧 Configuration Instructions
### TTS Voice Synthesis Configuration
1. **Obtain API Key**
- Visit the [ByteDance Speech Open Platform](https://openspeech.bytedance.com/).
- Register an account and create an application.
- Obtain AppID and access token.
2. **Configuration Steps**
- Enable voice notifications in the extension settings.
- Enter AppID and Token.
- Select voice type and speech rate.
- Test the voice synthesis functionality.
### AI Virtual Character Configuration
1. **Enable Character Assistant**
- Turn on "Character Assistant" in the popup settings.
- The character will appear on all webpages.
- Supports adaptive focus mode.
2. **Character Interaction Settings**
- Click the character to start a conversation.
- Supports text and voice input.
- Conversation history is automatically saved.
3. **MCP Function Configuration**
- The character supports intelligent task detection.
- Can directly call MCP services.
- Supports arxiv search, system control, etc.
### MCP Backend Service Configuration
1. **Install MCP Service**
```bash
# Navigate to MCP directory
cd MCP
# Install dependencies
cd native-host && npm install && cd ..
cd client && npm install && cd ..
# Configure Native Messaging
./scripts/install.sh
```
2. **Start MCP Service**
```bash
# Start all services
./scripts/start-all.sh
# Or start separately
cd native-host && npm start &
cd client && npm start &
```
3. **Verify MCP Connection**
- In the character dialogue, enter "play music" or "search paper."
- Check the MCP service logs.
- Confirm that the functionality is working properly.
### AI Assistant Configuration
1. **Basic Settings**
- Enable AI notification functionality.
- Configure API key (if needed).
- Set pre-generation time.
2. **Custom Prompt**
- Expand the "Prompt Settings" panel.
- Edit system prompts and user templates.
- Save settings and test the effects.
## 🚀 User Guide
### Basic Usage Flow
1. **Set Focus Time**
- Open the extension popup.
- Set the focus duration (recommended 25 minutes).
- Click "Start Focus."
2. **Configure Blocked Websites**
- Add distracting websites.
- Choose blocking mode (complete blocking/learning mode).
- Test the blocking effect.
3. **Enable AI Virtual Character**
- Turn on "Character Assistant" in the popup.
- The character will appear in the lower right corner of the webpage.
- Click the character to start a conversation.
4. **Personalization Settings**
- Configure notification methods (sound/TTS/AI).
- Adjust interface themes.
- Customize AI prompts.
### Virtual Character Usage
#### Basic Conversation
- **Text Chat**: Click the character and enter text in the dialogue box.
- **Voice Conversation**: Click the microphone icon to record a voice message.
- **Voice Reply**: The character will respond with TTS voice (requires TTS configuration).
#### MCP Intelligent Features
- **Academic Search**: "Help me search for papers related to machine learning."
- **System Control**: "Play music" / "Open application."
- **Chrome Search**: "Help me search for Python tutorials."
- **System Information**: "Show system information."
#### Character Animations
- **Idle Animation**: The character will automatically play idle animations.
- **Interaction Animation**: Displays thinking and speaking animations during conversations.
- **Focus Adaptation**: Character behavior will adjust accordingly in focus mode.
### Advanced Features
#### Learning Mode Website Configuration
```javascript
// Example of custom CSS selectors
{
"baidu.com": [
".result-op", // Hide promotional content
"#content_right", // Hide right-side ads
".c-recomm-wrap" // Hide recommended content
]
}
```
#### AI Prompt Customization Example
**System Prompt**:
```
You are a gentle learning partner, specifically helping users relax and rest after focused study.
Your replies should:
1. Be short and warm (no more than 50 words).
2. Be full of encouragement and care.
3. Suggest suitable rest activities.
4. Use a friendly and natural tone.
```
**User Template**:
```
The user has just completed {duration} minutes of focused study, and it is now {current time}.
Please generate a warm break reminder that acknowledges the user's efforts,
and suggests suitable ways to rest. The tone should be as natural and friendly as a caring friend.
```
## 🚀 Future Feature Plans
### 🔮 Planned Features
#### 1. 👁️ Visual Monitoring System
> **Status**: 🔄 In Planning | **Expected Release**: v1.3.0
**Core Features**:
- **Fatigue Detection** - Detects eye fatigue through the camera and intelligently reminds to rest.
- **Distraction Monitoring** - Detects when the user's attention is diverted and issues timely focus reminders.
- **Posture Monitoring** - Monitors sitting posture and head position, reminding to maintain correct posture.
- **Rest Reminders** - Personalized rest suggestions based on visual data.
**Technical Implementation**:
- Integrate WebRTC and MediaDevices API for camera access.
- Use TensorFlow.js for local machine learning inference.
- Eye tracking and facial expression recognition algorithms.
- Privacy-first design, with all data processed locally.
**Privacy Protection**:
- ✅ All visual data is processed locally and not uploaded to the cloud.
- ✅ Users have complete control over camera permissions.
- ✅ Optional feature, default off.
#### 2. 🎭 Dynamic Character Interaction System
> **Status**: 🔄 In Planning | **Expected Release**: v1.4.0
**Core Features**:
- **One-Click Character Generation** - AI-based customization of character appearance, personality, and voice.
- **Dynamic Expression System** - Rich facial expressions and emotional expressions.
- **3D Character Models** - Supports 3D virtual characters for a more realistic interaction experience.
- **Character Store** - Community-shared character templates and custom characters.
**Technical Implementation**:
- Integrate Three.js or Babylon.js for 3D rendering.
- Use Stable Diffusion API for character appearance generation.
- Personality and dialogue style customization based on LLM.
- Supports Live2D and VRM format character models.
**Customization Options**:
- 🎨 Appearance customization (hairstyle, clothing, accessories).
- 🗣️ Voice synthesis (tone, pitch, accent).
- 💭 Personality settings (lively, gentle, serious, humorous).
- 🎭 Professional roles (study partner, work assistant, fitness coach).
#### 3. 🤖 MCP-Powered AI Assistant
> **Status**: 🔄 In Planning | **Expected Release**: v1.5.0 | **Inspired by**: [Manus](https://github.com/manus-ai/manus)
**Core Features**:
- **Academic Research Assistant** - Deep integration with academic databases like arxiv, Google Scholar, PubMed, etc.
- **Code Development Assistant** - GitHub integration for code search, PR management, and issue tracking.
- **System Automation** - Local application control, file management, and system monitoring.
- **Workflow Automation** - Custom task chains and intelligent workflow orchestration.
**MCP Service Extensions**:
```
📚 Academic Research
├── arxiv-mcp-server # Paper search and download
├── scholar-mcp-server # Google Scholar integration
├── pubmed-mcp-server # Medical literature search
└── zotero-mcp-server # Reference management
💻 Development Tools
├── github-mcp-server # GitHub API integration
├── gitlab-mcp-server # GitLab support
├── docker-mcp-server # Container management
└── vscode-mcp-server # Editor control
🖥️ System Control
├── macos-mcp-server # macOS system control
├── windows-mcp-server # Windows system control
├── linux-mcp-server # Linux system control
└── browser-mcp-server # Browser automation
```
**Intelligent Features**:
- 🧠 **Intelligent Task Understanding** - Natural language converted into specific operational steps.
- 🔄 **Automated Workflows** - Automated execution of complex tasks.
- 📊 **Work Analysis** - Work efficiency analysis and optimization suggestions.
- 🎯 **Goal Management** - Long-term goal tracking and progress management.
**Security Features**:
- 🔒 Permission control and operation confirmation mechanisms.
- 🛡️ Sandbox environment for executing sensitive operations.
- 📝 Detailed operation logs and audits.
- ⚠️ Multiple confirmations for dangerous operations.
#### 4. 🎭 AI Learning Companion System
> **Status**: 🔄 In Planning | **Expected Release**: v1.6.0 | **Core Technology**: Long Cache Agent Architecture
**Core Features**:
- **🏃♂️ Keep Fitness Coach** - Personalized workout plans, movement guidance, progress tracking.
- **👨🍳 Culinary Master Assistant** - Recipe recommendations, cooking guidance, nutritional pairing.
- **🗣️ Language Learning Partner** - Multilingual conversation practice, pronunciation correction, cultural exchange.
- Chinese context friends (dialects, Mandarin, classical poetry).
- English context friends (British, American, business English).
- Japanese context friends (keigo, Kansai dialect, anime language).
- **📚 Subject Professional Tutor** - Guidance in subjects like mathematics, physics, programming, design, etc.
**Long Cache Agent Technology Architecture**:
```
🧠 Agent Core Layer
├── Long-Term Memory # Long-term memory storage (user preferences, learning history).
├── Context Cache # Context cache (session state, task progress).
├── Personality Engine # Personality engine (character consistency, emotional state).
└── Learning Adapter # Learning adapter (user behavior analysis, personalized adjustments).
🎭 Character Specialization Layer
├── Fitness Coach # Fitness coach (workout plans, movement library, health data).
├── Culinary Master # Culinary master (recipe library, cooking skills, nutritional knowledge).
├── Language Partner # Language partner (multilingual models, cultural background, conversation scenarios).
└── Subject Tutor # Subject tutor (knowledge base, teaching methods, progress tracking).
🔄 Persistent Storage Layer
├── User Profile Store # User profile storage.
├── Conversation History # Conversation history (by character, by topic).
├── Progress Tracking # Progress tracking (learning curve, achievement system).
└── Preference Learning # Preference learning (automatic optimization, personalized recommendations).
```
**Learning Companion Intelligent Features**:
- 🎭 **Character Consistency** - Each companion maintains a unique personality and professional characteristics.
- 📚 **Progressive Learning** - Automatically adjusts difficulty and content based on user level.
- 🔄 **Habit Formation** - Long-term companionship helps users establish healthy learning and living habits.
- 💡 **Creative Inspiration** - Provides personalized creative suggestions based on user interests and history.
- 🎯 **Goal-Oriented** - Helps users set and achieve specific learning or life goals.
**Long Cache Agent Core Advantages**:
- 🧠 **Deep Memory** - Remembers every interaction with the user, forming a complete user profile.
- 🎨 **Personalized Evolution** - Becomes more understanding of the user as interactions increase.
- 🔄 **Continuous Experience** - Seamless conversation experience across devices and time.
- 📈 **Growth Trajectory** - Records and analyzes the user's learning growth process.
**MCP Service Extensions**:
```
🎭 Learning Companion Services
├── fitness-mcp-server # Fitness data and plan management.
├── recipe-mcp-server # Recipe and nutrition database.
├── language-mcp-server # Multilingual learning resources.
└── education-mcp-server # Educational resources and progress tracking.
```
### 🎯 Release Roadmap
> **Note that the order may not be fixed!**
| Version | Feature | Status | Expected Time |
|---------|---------|--------|---------------|
| v1.2.1 | ✅ Basic Version of Virtual Character | Released | 2025-01 |
| v1.3.0 | 👁️ Visual Monitoring System | In Planning | 2025-03 |
| v1.4.0 | 🎭 Dynamic Character System | In Planning | 2025-05 |
| v1.5.0 | 🤖 MCP Intelligent Assistant | In Planning | 2025-07 |
| v1.6.0 | 🎭 AI Learning Companion System | In Planning | 2025-09 |
| v2.0.0 | 🌟 Complete Ecosystem | In Planning | 2025-11 |
### 🎭 Learning Companion Detailed Planning
#### Fitness Coach Companion 🏃♂️
**Personality Traits**: Sunny, positive, professional, patient.
**Core Features**:
- Personalized workout plan creation (based on user physique, goals, time).
- Real-time movement guidance and correction (combined with visual monitoring).
- Tracking and analysis of exercise data (heart rate, calories, progress curve).
- Nutritional advice and meal pairing.
- Motivation and habit formation (check-in system, achievement unlocking).
**Long Cache Features**:
- Remembers user exercise preferences and physical conditions.
- Tracks long-term fitness progress and goal achievement.
- Adjusts training intensity and methods based on user feedback.
- Learns user time habits and intelligently reminds to exercise.
#### Culinary Master Companion 👨🍳
**Personality Traits**: Passionate, creative, meticulous, foodie temperament.
**Core Features**:
- Recipe recommendations based on available ingredients.
- Step-by-step cooking guidance (text + voice).
- Nutritional pairing and healthy eating advice.
- Exploration of different cuisines and flavors.
- Teaching cooking skills and advanced guidance.
**Long Cache Features**:
- Remembers user taste preferences and dietary restrictions.
- Tracks user cooking progress and mastered dishes.
- Recommends seasonal and festive recipes.
- Learns user cooking habits and time arrangements.
#### Language Learning Partner 🗣️
**Personality Traits**: Friendly, knowledgeable, culturally sensitive, encouraging.
**Core Features**:
- Daily conversation practice in multiple languages.
- Pronunciation correction and intonation guidance.
- Introduction to cultural backgrounds and customs.
- Grammar explanations and vocabulary expansion.
- Scenario simulation and role-playing.
**Long Cache Features**:
- Remembers user language levels and learning progress.
- Tracks vocabulary mastery and grammar weaknesses.
- Selects conversation topics based on user interests.
- Adapts to user learning pace and style preferences.
#### Subject Professional Tutor 📚
**Personality Traits**: Rigorous, patient, heuristic, tailored teaching.
**Core Features**:
- Personalized learning plan creation.
- Concept explanations and knowledge point organization.
- Practice question recommendations and answer guidance.
- Teaching methods and techniques.
- Learning progress tracking and assessment.
**Long Cache Features**:
- Remembers user knowledge base and learning abilities.
- Tracks learning progress and mastery levels.
- Identifies user learning blind spots and difficulties.
- Adjusts teaching methods based on user feedback.
### 🔧 Long Cache Agent Technical Implementation
#### Core Technology Stack
```typescript
// Long-term memory storage architecture
interface LongTermMemory {
userProfile: UserProfile; // User profile
conversationHistory: ConversationHistory[]; // Conversation history
learningProgress: LearningProgress; // Learning progress
preferences: UserPreferences; // User preferences
}
// Context cache system
interface ContextCache {
sessionState: SessionState; // Session state
taskProgress: TaskProgress[]; // Task progress
emotionalState: EmotionalState; // Emotional state
recentInteractions: Interaction[]; // Recent interactions
}
// Personalization adaptation engine
interface PersonalityEngine {
characterTraits: CharacterTraits; // Character traits
responseStyle: ResponseStyle; // Response style
teachingMethod: TeachingMethod; // Teaching method
adaptationRules: AdaptationRule[]; // Adaptation rules
}
```
#### Data Persistence Strategy
- **Local Storage**: Uses IndexedDB for large-capacity data storage.
- **Cloud Sync**: Optional encrypted cloud backup (user-controlled).
- **Incremental Updates**: Only syncs changed data to reduce bandwidth consumption.
- **Version Control**: Supports data rollback and historical version viewing.
#### Privacy Protection Mechanisms
- **End-to-End Encryption**: All user data is stored encrypted locally.
- **Anonymization Processing**: Sensitive information is automatically anonymized.
- **User Control**: Complete data control, with the ability to delete at any time.
- **Transparency**: Clear data usage instructions and privacy policies.
## 📈 Update History
### v1.2.1 (2025-06-11) 🎉
- **Major Update**: Bug fixes.
- **New**: Voice input and TTS voice replies.
- **Optimization**: User interface and interaction experience.
- **Improvement**: Interface responsiveness.
- **Fixes**: Multiple stability issues and memory leaks.
### v1.2.0 (2025-04-15) 🎉
- **Major Update**: Comprehensive refactoring for more stable functionality and interaction.
- **New**: AI virtual character system, supporting voice dialogue and animations.
- **New**: MCP architecture integration, supporting arxiv search and system control.
- **New**: TTS voice synthesis functionality, supporting ByteDance API.
- **New**: AI personalized notification system and custom prompts.
- **Optimization**: User interface and interaction experience, supporting dark mode.
- **Fixes**: Multiple stability issues and memory leaks.
### v1.0.0 (2025-02-21)
- **Fixes**: Critical bug fixes.
- **Optimization**: Improved notification experience.
- **Improvement**: Interface responsiveness.
### v0.0.1 (2025-02-05) 🌱
- **Initial Release**: Basic focus time management functionality.
- **Core Features**: Website blocking, time timing, notification reminders.
## 🤝 Contribution Guidelines
We welcome community contributions! Please follow these steps:
1. **Fork the project** and create a feature branch.
2. **Write code** and ensure it passes tests.
3. **Submit a PR** with a detailed description of the changes.
4. **Code review** will be merged into the main branch after approval.
### Development Standards
- Use TypeScript for type-safe development.
- Follow ESLint and Prettier code standards.
- Write unit tests to cover new features.
- Update relevant documentation.
## 📄 License
This project is licensed under the [MIT License](LICENSE).
## 🙏 Acknowledgments
- [Chrome Extension Boilerplate](https://github.com/Jonghakseo/chrome-extension-boilerplate-react-vite) - Project scaffold.
- [ByteDance TTS](https://openspeech.bytedance.com/) - Voice synthesis service.
- [deepseek](https://www.deepseek.com/) - AI model service.
- [MCP (Model Context Protocol)](https://github.com/modelcontextprotocol) - Intelligent assistant architecture.
- [arxiv-mcp-server](https://github.com/blazickjp/arxiv-mcp-server) - Academic search service.
- [Manus](https://github.com/manus-ai/manus) - Inspiration for MCP intelligent assistant.
- All contributors and users for their support.
## 💝 Support the Project
If this project has been helpful to you, feel free to:
- ⭐ Give the project a star.
- 🐛 Report bugs and suggestions.
- 🔀 Submit pull requests.
- 📢 Recommend it to friends.
> **Open Source Commitment**: This project is committed to being permanently open-source and free, adhering to the principle of transparent development! ✅🍃🎉
---
<div align="center">
**Do not think of good or evil, just at this moment, which is the true face of the original seat.** 🧘♂️
*Focus on the present, live in the moment.*
</div>
Connection Info
You Might Also Like
markitdown
Python tool for converting files and office documents to Markdown.
everything-claude-code
Complete Claude Code configuration collection - agents, skills, hooks,...
awesome-claude-skills
A curated list of awesome Claude Skills, resources, and tools for...
skyflo
Self-hosted AI Agent for Kubernetes & DevOps
memo
Memo MCP -- save and restore conversation across agents
WeChat-MCP
WeChat MCP Server automates WeChat on macOS, enabling LLMs to interact with chats.