MCP-Server

cr0w1234/MCP-Server

3.1

If you are the rightful owner of MCP-Server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

The Model Context Protocol (MCP) server is a robust tool designed for database querying and external API integration, optimized for deployment on Google Cloud Run.

Tools
5
Resources
0
Prompts
0

MCP Server and RAG UI Documentation

This repository contains two main components:

  1. working-mcp-on-cloudrun: MCP server implementation and deployment tools
  2. rag-mcp-server: Simple web UI for interacting with MCP tools

📖 Quick Start: See for detailed setup instructions.

📁 Project Structure

test-rag-mcp/
├── working-mcp-on-cloudrun/          # MCP server implementation
│   ├── readonly-db-mcp-server.py     # Read-only database MCP server
│   ├── test_mcp_server.py            # MCP server testing utilities
│   ├── test_openai.py                # OpenAI API integration tests
│   ├── pyproject.toml                # Python dependencies
│   ├── uv.lock                       # Dependency lock file
│   └── Dockerfile                    # Cloud Run deployment
└── rag-mcp-server/                   # Web UI for MCP interaction
    ├── ui.py                         # Flask web application
    ├── rag_system.py                 # RAG (Retrieval-Augmented Generation) system
    └── documents/                    # Document storage for RAG

🚀 MCP Server (working-mcp-on-cloudrun)

Overview

The MCP (Model Context Protocol) server provides tools for database querying and external API integration. It's designed to be deployed on Google Cloud Run for scalable, serverless operation.

Key Files

readonly-db-mcp-server.py

Purpose: Main MCP server implementation with read-only database access and TMDB API integration.

Features:

  • Read-only Database Queries: Secure PostgreSQL access with SELECT-only restrictions
  • TMDB API Integration: Intelligent movie/TV show data retrieval
  • Connection Pooling: Robust database connection management
  • Security Validation: SQL injection protection and query validation

Available Tools:

  • query_demo_db(sql): Execute read-only SQL queries
  • tmdb_intelligent_call(request): Smart TMDB API calls
  • http_get(url, headers): Generic HTTP GET requests
  • add(a, b), subtract(a, b): Basic arithmetic operations
test_mcp_server.py

Purpose: Testing utilities for MCP server functionality.

Usage:

# Test MCP tools connectivity locally
uv run test_mcp_server.py
test_openai.py

Purpose: OpenAI API integration testing and validation.

Features:

  • Tests OpenAI client configuration
  • Validates API key authentication
  • Tests MCP tool integration with OpenAI

Deployment

Local Development
cd working-mcp-on-cloudrun
pip install -r requirements.txt
python readonly-db-mcp-server.py
Cloud Run MCP Server Deployment

Follow this guide: https://cloud.google.com/run/docs/tutorials/deploy-remote-mcp-server#source

Environment Variables
DATABASE_URL=postgresql://user:pass@host:port/db
MCP_SERVER_URL=https://your-mcp-server-url.run.app/mcp/
TMDB_API_KEY=your_tmdb_api_key
TMDB_ACCOUNT_ID=your_account_id
TMDB_SESSION_ID=your_session_id
OPENAI_API_KEY=your_openai_api_key

🌐 RAG UI (rag-mcp-server)

Overview

A simple Flask-based web interface that combines traditional RAG (Retrieval-Augmented Generation) with MCP tool integration for enhanced query capabilities.

Key Files

ui.py

Purpose: Main Flask web application providing the user interface.

Features:

  • Dual Search Modes: Traditional RAG and MCP-powered search
  • Conversation Interface: Scrollable chat-like interface with conversation history
  • Document References: Links to source documents and references
  • MCP Integration: Seamless integration with MCP tools

Endpoints:

  • GET /: Main web interface
  • POST /ask: Traditional RAG search
  • POST /ask_mcp: MCP-powered database search
  • GET /doc/<path>: Document serving

Conversation Memory:

# Global conversation state management
current_response_id = None

@app.route('/ask_mcp', methods=['POST'])
def ask_mcp():
    global current_response_id
    # Maintains conversation continuity across requests
rag_system.py

Purpose: RAG (Retrieval-Augmented Generation) system implementation.

Features:

  • Document Processing: Handles PDF, DOCX, and other document formats
  • Vector Embeddings: Creates and manages document embeddings
  • Semantic Search: Retrieves relevant document chunks
  • Answer Generation: Generates answers based on retrieved context

Usage:

from rag_system import RAGService

rag_service = RAGService()
result = rag_service.answer_question("Your question here")

UI Features

Conversation Interface
  • Scrollable Chat: Full conversation history with automatic scrolling
  • Message Types: Distinct styling for user questions and assistant responses
  • Markdown Support: Rich text formatting for responses
  • Reference Links: Clickable links to source documents
Search Modes
  1. Traditional RAG (/ask): Document-based search with vector embeddings
  2. MCP Search (/ask_mcp): Database querying with conversation memory

🔧 Setup and Installation

Prerequisites

  • Python 3.8+
  • PostgreSQL database
  • OpenAI API key
  • TMDB API key (optional)
  • Google Cloud account (for deployment)

Local Setup

  1. Run the Applications
# Terminal 1: Start MCP server (no need if using an already deployed MCP server url)
cd working-mcp-on-cloudrun
python readonly-db-mcp-server.py

# Terminal 2: Start RAG UI
cd rag-mcp-server
python ui.py

🧪 Testing

MCP Server Testing

cd working-mcp-on-cloudrun

# Test database connectivity
python test_mcp_server.py

# Test OpenAI integration
python test_openai.py

RAG UI Testing

cd rag-mcp-server

# Start the UI
python ui.py

# Access at http://localhost:5000
# Test both search modes:
# 1. Traditional RAG: Ask questions about documents
# 2. MCP Search: Query the database with natural language

🔒 Security Considerations

Database Security

  • Read-only Access: Only SELECT queries are allowed
  • SQL Injection Protection: Comprehensive query validation
  • Connection Pooling: Secure connection management
  • Environment Variables: Sensitive data stored in environment variables

API Security

  • Rate Limiting: Implement rate limiting for production
  • Authentication: Add authentication for production deployments
  • CORS: Configure CORS policies appropriately

📊 Monitoring and Logging

Google Cloud Logging

The MCP server includes comprehensive logging for Google Cloud:

logger.info(f"MCP Query Request: {question}")
logger.info(f"Tool calls executed: {len(resp.tool_calls)}")

Log Levels

  • INFO: Normal operation logs
  • WARNING: Connection pool issues
  • ERROR: Database connection failures

🤝 Contributing

  1. Fork the repository
  2. Create a feature branch
  3. Make your changes
  4. Add tests for new functionality
  5. Submit a pull request

🆘 Troubleshooting

Common Issues

  1. Database Connection Errors

    • Verify DATABASE_URL format
    • Check database user permissions
    • Ensure database is accessible

  2. MCP Server Not Starting

    • Check environment variables
    • Verify port availability
    • Check dependency installation

  3. RAG UI Not Loading

    • Ensure Flask dependencies are installed
    • Check port 5000 availability
    • Verify document directory exists