Content
# Qubase MCP
A powerful, cross-platform AI Chat Client implementing the Model Context Protocol (MCP) for seamless AI interactions.
---
## Overview
Qubase MCP is a sophisticated chat interface that revolutionizes AI interactions on desktop and mobile devices. Built with Flutter and implementing the Model Context Protocol, it provides a unified interface for multiple Large Language Models (LLMs) while ensuring secure, efficient, and context-aware conversations.
## Features
### Core Capabilities
| Feature | Description |
|---------|-------------|
| Universal Compatibility | Works seamlessly across all major platforms |
| Model Flexibility | Connect to any supported LLM without changing workflow |
| Context Awareness | Leverages MCP for maintaining conversation context |
| Enterprise Ready | Built with security and scalability in mind |
### Platform Support
| Platform | Status |
|----------|---------|
| Desktop (macOS, Windows, Linux) | Available |
| Mobile (iOS, Android) | Available |
| Web | Coming Soon |
### Supported AI Models
- OpenAI (GPT-3.5, GPT-4)
- Anthropic Claude
- OLLama (Local Models)
- DeepSeek
- Custom Model Support
---
## Local LLM Setup
### Android Setup
1. **Install Termux**
- Download Termux ARM64 V8 from [Termux GitHub](https://github.com/termux/termux-app/releases)
- Install and open Termux
- Run `termux-setup-storage` to grant storage permissions
- Run `termux-change-repo` to select package mirror
- Update with `pkg upgrade`
2. **Required Packages**
```bash
# Install Tur repository
pkg install tur-repo
# Install Ollama and Zellij
pkg install ollama
pkg install zellij
```
3. **Android Configuration**
- Enable Developer Options: Settings > About device > Tap "Build number" 7 times
- In Developer options, enable "Disable child process restrictions"
4. **Model Operations**
```bash
# Start Ollama server
ollama serve
# In a new terminal, run models:
ollama run deepseek-r1.5b # For DeepSeek
ollama run llama3.2 # For Llama3
```
5. **Control Commands**
| Action | Command |
|--------|---------|
| Stop output | `CTRL + C` |
| Exit model | `CTRL + D` |
| Clear screen | `CTRL + L` |
| Stop server | `ps aux \| grep ollama` then `kill [PID]` |
### Desktop Setup
_Coming soon..._
---
## System Requirements
### Hardware Requirements
| Component | Minimum | Recommended |
|-----------|---------|-------------|
| RAM | 4GB | 8GB |
| Storage | 2GB free | 4GB free |
| Processor | Intel/AMD x64 or ARM64 | Modern multi-core |
### Software Requirements
| Component | Version |
|-----------|---------|
| Windows | 10 or later |
| macOS | 10.15 or later |
| Ubuntu | 20.04 or later |
| Flutter SDK | 3.0 or later |
| Git | Latest stable |
---
## Installation Guide
### Prerequisites
1. **Flutter Setup**
```bash
git clone https://github.com/flutter/flutter.git
export PATH="$PATH:`pwd`/flutter/bin"
flutter doctor
```
2. **System Dependencies**
```bash
# Linux
sudo apt-get update
sudo apt-get install libsqlite3-0 libsqlite3-dev
# macOS
brew install sqlite3
# Windows
# SQLite included in Flutter Windows setup
```
3. **Development Tools**
```bash
# Using uv (Recommended)
brew install uv
# Alternative: npm
brew install node
```
### Application Setup
1. **Installation**
```bash
git clone https://github.com/qubasehq/Qmcp.git
cd Qmcp
flutter pub get
```
2. **Launch**
```bash
# Desktop platforms
flutter run -d <platform> # macos, windows, linux
# Mobile development
flutter run -d <device-id>
```
---
## Configuration
### Initial Setup
1. **API Configuration**
- Launch Qubase MCP
- Navigate to Settings > API Configuration
- Enter LLM API credentials
- Configure custom endpoints (if needed)
2. **MCP Server Setup**
- Access Settings > MCP Server
- Choose installation method
- Configure server settings
- Select default AI model
### File Locations
| Purpose | Path |
|---------|------|
| Configuration | `~/Library/Application Support/qubase_mcp/mcp_server.json` |
| Logs | `~/Library/Application Support/run.daodao.qubase_mcp/logs` |
| Application Data | `~/Library/Application Support/qubase_mcp` |
### Reset Application
```bash
# Clear all data (use with caution)
rm -rf ~/Library/Application\ Support/run.daodao.qubase_mcp
rm -rf ~/Library/Application\ Support/qubase_mcp
```
---
## Troubleshooting
| Issue | Solution |
|-------|----------|
| Connection Issues | Verify API keys and network connectivity |
| Performance Problems | Check system resources and clear cache |
| Model Errors | Validate model configurations and quotas |
For additional support, visit our [Issues](https://github.com/qubasehq/Qmcp/issues) page.
---
## Development Roadmap
### Upcoming Features
| Feature | Description |
|---------|-------------|
| MCP Server Marketplace | Easy discovery and deployment of community servers |
| Enhanced Integration | Automated server setup and cloud sync support |
| RAG Implementation | Document processing and knowledge base integration |
| UI/UX Improvements | Custom themes, keyboard shortcuts, mobile optimization |
---
## Contributing
1. Fork the repository
2. Create a feature branch
3. Commit your changes
4. Push to your branch
5. Open a Pull Request
---
## Acknowledgments
- Model Context Protocol (MCP) Team
- MCP CLI Contributors
- Flutter Community
- Open Source AI Community
---
## License
Licensed under Apache License 2.0 - see the [LICENSE](./LICENSE) file for details.
