🐉 Dragon - Multi-Provider AI CLI

A powerful command-line AI tool supporting multiple providers, inspired by Claude Code.

✨ Features

• 🔄 Multi-Provider Support: Anthropic Claude, OpenAI, Google Gemini, DeepSeek, and Chinese LLMs (Qwen, Moonshot, Yi, Doubao)
• 💬 REPL Interface: Interactive chat with streaming responses
• 🛠️ Tool System: Execute commands, read/write files, search code, web fetch, and more
• 🤖 Agent Support: Spawn sub-agents for complex tasks
• 🔐 Encryption: AES-256-GCM encryption for API keys and sensitive fields
• ⚡ Performance Monitoring: Built-in timing and cost tracking
• ⚙️ Configurable: Easy configuration for multiple API keys

📦 Installation

npm

npm install -g opendragon

# Run
dragon

Build from source

git clone https://github.com/zhaojiewen/open-dragon.git
cd open-dragon
npm install
npm run build
npm link

🚀 Quick Start

# 1. Initialize config (with optional encryption)
dragon init
dragon init --encrypt          # Encrypt API keys with a password

# 2. Edit config to add your API keys
dragon config edit
# Or manually: vim ~/.dragon/config.json

# 3. Start interactive chat
dragon

# Or specify a provider and model
dragon chat --provider openai --model gpt-4o

⚙️ Configuration

Config file: ~/.dragon/config.json

{
  "defaultProvider": "anthropic",
  "providers": {
    "openai": {
      "apiKey": "YOUR_OPENAI_API_KEY",
      "models": ["gpt-4o", "gpt-4-turbo"],
      "defaultModel": "gpt-4o"
    },
    "anthropic": {
      "apiKey": "YOUR_ANTHROPIC_API_KEY",
      "models": ["claude-opus-4-7", "claude-sonnet-4-6", "claude-haiku-4-5"],
      "defaultModel": "claude-opus-4-7"
    },
    "gemini": {
      "apiKey": "YOUR_GEMINI_API_KEY",
      "models": ["gemini-1.5-pro", "gemini-1.5-flash"],
      "defaultModel": "gemini-1.5-pro"
    },
    "deepseek": {
      "apiKey": "YOUR_DEEPSEEK_API_KEY",
      "models": ["deepseek-chat", "deepseek-reasoner"],
      "defaultModel": "deepseek-chat"
    },
    "qwen": {
      "apiKey": "YOUR_QWEN_API_KEY",
      "baseUrl": "https://dashscope.aliyuncs.com/compatible-mode/v1",
      "models": ["qwen-max", "qwen-plus", "qwen-turbo"],
      "defaultModel": "qwen-max"
    }
  },
  "tools": {
    "enabled": ["bash", "read", "write", "edit", "glob", "grep", "webfetch", "websearch", "agent"]
  },
  "workspace": {
    "paths": [],
    "writeEnabled": true,
    "enforceBounds": false,
    "allowHomeDir": true
  },
  "logging": {
    "level": "info",
    "logFile": "/path/to/dragon.log",
    "enableConsole": true
  }
}

Workspace Configuration

Restrict file operations to specific directories:

{
  "workspace": {
    "paths": ["/home/user/projects"],
    "writeEnabled": true,
    "enforceBounds": true,
    "allowHomeDir": true
  }
}

📝 CLI Commands

| Command | Description | |---------|-------------| | dragon | Start REPL (default) | | dragon init | Initialize config file | | dragon init --encrypt | Initialize with encrypted API keys | | dragon config | Show config file location | | dragon config edit | Open config in editor | | dragon config show | Print config contents | | dragon config validate | Validate configuration | | dragon chat | Start interactive chat | | dragon chat -p <provider> -m <model> | Chat with specific provider/model | | dragon --monitor | Start with performance monitoring | | dragon --help | Show help | | dragon --version | Show version |

💻 REPL Commands

| Command | Description | |---------|-------------| | /help | Show available commands | | /clear | Clear conversation history | | /history | Show conversation history | | /history save <name> | Save history to file | | /history load <name> | Load history from file | | /provider [name] | Show or switch provider | | /model [name] | Show or switch model | | /tools | List available tools | | /auto [out] | Toggle auto-approve dangerous tools | | /ask | Require confirmation for all tools | | /workspace <path> | Set workspace directory | | /save-tokens | Save token usage data | | /cost | Show token usage & cost estimate | | /perf | Show performance report (needs --monitor) | | /debug [on|off] | Toggle debug mode | | /exit, /quit | Exit REPL |

🔧 Available Tools

| Tool | Description | |------|-------------| | bash | Execute shell commands | | read | Read file contents | | write | Write to files | | edit | Edit files with string replacement | | glob | Find files by pattern | | grep | Search file contents | | webfetch | Fetch web page content | | websearch | Web search (requires API config) | | agent | Spawn sub-agent for complex tasks |

🌐 Supported Providers

| Provider | Models | |----------|--------| | Anthropic | claude-opus-4-7, claude-sonnet-4-6, claude-haiku-4-5 | | OpenAI | gpt-4o, gpt-4-turbo, gpt-3.5-turbo | | Google Gemini | gemini-1.5-pro, gemini-1.5-flash | | DeepSeek | deepseek-chat, deepseek-reasoner | | Qwen (通义千问) | qwen-max, qwen-plus, qwen-turbo, qwen-long | | Moonshot (月之暗面) | moonshot-v1-8k, moonshot-v1-32k, moonshot-v1-128k | | Yi (零一万物) | yi-lightning, yi-large, yi-medium | | Doubao (豆包) | doubao-pro-4k, doubao-pro-32k, doubao-pro-128k |

➕ Custom Providers

Any OpenAI-compatible API can be added:

{
  "providers": {
    "my-custom-llm": {
      "apiKey": "YOUR_API_KEY",
      "baseUrl": "https://api.example.com/v1",
      "models": ["model-1", "model-2"],
      "defaultModel": "model-1"
    }
  }
}

🔐 Security

API Key Encryption

AES-256-GCM encryption for sensitive config fields (apiKey, token, secret, password):

# Initialize with encryption
dragon init --encrypt

# Run with encrypted config
DRAGON_PASSWORD=yourpassword dragon

# Or use the --password flag (visible in shell history - less secure)
dragon init --encrypt --password yourpassword

Security features:

• AES-256-GCM with random IV per encryption
• PBKDF2 key derivation (100,000 iterations, SHA-512)
• Automatic detection of sensitive fields
• Secure file permissions (0o600)
• Timing-safe comparison for key verification

📊 Performance Monitoring

# Enable via flag
dragon --monitor

# Or via environment variable
DRAGON_PERF_MONITOR=true dragon

# View report in REPL
> /perf

Tracked metrics: API call times, tool execution stats, config load time, streaming performance.

🐛 Debug Mode

# Environment variable
DRAGON_DEBUG=true dragon

# REPL command
> /debug on
> /debug off

# Log levels via env
DRAGON_LOG_LEVEL=0  # DEBUG
DRAGON_LOG_LEVEL=1  # INFO (default)
DRAGON_LOG_LEVEL=2  # WARN
DRAGON_LOG_LEVEL=3  # ERROR
DRAGON_LOG_LEVEL=4  # SILENT

📋 Development

npm install          # Install dependencies
npm run build        # Compile TypeScript
npm run dev          # Run directly with tsx
npm run bundle       # Create single-file dist
npm run lint         # Run ESLint

🧪 Testing

npm test              # Run all tests
npm run test:watch    # Watch mode
npm run test:coverage # With coverage report (text + coverage/index.html)
npx vitest tests/unit/path/to/test.test.ts  # Run single test file

Test structure:

tests/
├── unit/
│   ├── tools/        # Tool tests (bash, read, write, edit, glob, grep, registry)
│   ├── config/       # Config tests
│   ├── encryption/   # Encryption tests
│   ├── providers/    # Provider tests
│   └── utils/        # Error, logger, performance tests

🔄 CI/CD

GitHub Actions workflows (.github/workflows/):

• CI: Runs on push to main, tests on Node.js 18.x/20.x/22.x, generates coverage reports
• Release: Triggers on GitHub release, publishes to npm, uploads bundle

Required secrets: NPM_TOKEN, CODECOV_TOKEN (optional)

📄 License

MIT

🤝 Contributing

Issues and pull requests welcome!

📮 Links

• GitHub Repository
• npm Package
• Report Bug