mcp-system-infra

🚀 Intelligent Architecture Recommendation Engine: Tailored for Your System

In today's rapidly evolving digital landscape, how can you quickly and efficiently build a scalable and reliable technical infrastructure? The Intelligent Architecture Recommendation Engine is here to solve that challenge.

Based on key parameters—QPS (queries per second), concurrent users, daily active users, business type, database choice, and AI model size—this tool automatically generates:

💡 Optimal server resource allocation
🧩 Required middleware module combinations
🏗️ Recommended overall system architecture
☁️ Suggested cloud providers and deployment strategies
📊 One-click export of Markdown report + architecture diagram

✨ Key Benefits

✅ Fully Parameter-Driven, Business-Oriented

Simply provide the following parameters:

--qps: Peak request throughput
--concurrentUsers: Number of concurrent connections
--uad: Daily Active Users (UAD)
--type: Business type (web / ai)
--db: Database type (relational / nosql / analytics)
--model: AI model size (small / medium / large)

The system will automatically assess and recommend:

CPU / Memory / Network configuration
Redis cache capacity and eviction strategy
Message queue type and concurrency handling
Whether to adopt a microservices architecture
Whether to enable distributed architecture and GPU inference clusters

🗺️ Architecture Recommendation Diagram

The system automatically generates a Mermaid diagram to clearly represent component relationships:

flowchart TD User[User Request] --> Nginx[Nginx Load Balancer] Nginx --> Service[Main Business Service Node] Service --> DB[Database] Service --> Redis[Redis Cache] Service --> MQ[Message Queue] Service --> GPU[AI Inference GPU Node] MQ --> Consumer[Asynchronous Consumer]

▶️ Quick Start

CLI

npx -y mcp-system-infra

MCP Server Configuration

{ "mcpServers": { "mcp-system-infra": { "command": "npx", "args": [ "-y", "mcp-system-infra" ] } } }

MCP Example:

Please help design a web-based system architecture report with the following specifications:

QPS (Queries Per Second): 100
Concurrent Users: 50
Daily Active Users: 300
Database Type: Relational
Model Size: Medium

💭 Murmurs

This project is for educational and internal use only. Contributions and feedback are welcome.
For feature customization, web deployment, or enterprise integration, please contact the project maintainer.

Business Contact Email: deeppathai@outlook.com

🧠 MCP Access Addresses

🌐 ModelScope MCP Address
For testing and integrating mcp-system-infra directly within the ModelScope platform.
🛠️ Smithery.ai MCP Address
For visual configuration and invocation of the mcp-system-infra service via Smithery.