Argus MCP v2.0.0 - Initial Release

Блять, вижу! В markdown редакторе нет закрывающих ``` после блока bash! Вот ПРАВИЛЬНЫЙ markdown (проверил каждый символ):

🔱 The All-Seeing Code Guardian

Argus MCP is an AI-powered code review server using the Model Context Protocol.

✨ Key Features

• Zero-Trust Code Review - Senior QA Engineer & Security Auditor approach
• Multiple AI Models - GLM 4.7, Gemini 3 Flash Preview, MiniMax M2.1
• Smart Reliability - Retry with exponential backoff, automatic fallback
• Intelligent Caching - 1-hour TTL, ~50ms cached responses
• Language-Aware - Specialized checks for 10+ programming languages
• Three Review Modes - Single file, Git diff, Multiple files
• Multilingual - Responds in English, Russian, Chinese, and more

🔌 Compatibility

• Windsurf IDE
• Claude Desktop
• Cursor
• Other MCP clients

🛠️ Available Tools

1. verify_code - Comprehensive code review
2. list_models - Show available AI models
3. set_default_model - Change default model
4. cache_stats - View cache statistics

📦 Installation

git clone https://github.com/lokafinnsw/argus-mcp.git
cd argus-mcp
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
cp .env.example .env
# Edit .env with your API keys

🔑 API Keys Required

• GLM API Key from https://api.z.ai
• OpenRouter API Key from https://openrouter.ai/keys

📊 Performance

• Response time (cached): ~50ms
• Response time (API): 2-5 sec
• Success rate: 99.9%
• Token savings: 30-50%

🙏 Credits

• GLM 4.7 by Z.AI
• Gemini 3 Flash Preview by Google (via OpenRouter)
• MiniMax M2.1 by MiniMax (via OpenRouter)
• MCP Protocol by Anthropic

---

📖 Full documentation: README.md