MCP Orchestration System

# 🤖 MCP Agent System - User Guide ## 🎯 **How to Use Your User-Friendly MCP System** Your MCP (Model Context Protocol) system is now fully set up with MongoDB integration and multiple user-friendly interfaces. Here's how to use it: --- ## 🚀 **Quick Start** ### **1. Start the System** ```bash python production_mcp_server.py ``` ### **2. Choose Your Interface** #### **🌐 Web Interface (Recommended)** - Open: http://localhost:8000 - Features: Beautiful UI, real-time responses, query history - Best for: Interactive exploration and testing #### **💻 Interactive Command Line** ```bash python user_friendly_interface.py ``` - Features: Terminal-based chat interface - Best for: Power users and automation #### **⚡ Quick Single Queries** ```bash python quick_query.py "Your question here" ``` - Features: One-shot queries with instant results - Best for: Scripts and quick tests --- ## 💬 **What You Can Ask** ### **🔢 Math Calculations** ``` Calculate 25 * 4 What is 100 + 50? Compute 20% of 500 Solve 15 + 25 * 2 Find the square root of 144 ``` ### **🌤️ Weather Queries** ``` What is the weather in Mumbai? Mumbai weather Temperature in Delhi Weather forecast for Bangalore Climate in New York ``` ### **📄 Document Analysis** ``` Analyze this text: Your text content here Process document content Extract information from text Summarize this paragraph: Your content ``` --- ## 🌐 **Web Interface Guide** ### **Features:** - **Real-time Status**: See server, MongoDB, and agent status - **Query Input**: Type questions naturally - **Example Buttons**: Click to try sample queries - **Response Display**: Formatted results with all details - **Query History**: Track your previous questions - **Clear/Refresh**: Reset interface and update status ### **How to Use:** 1. Open http://localhost:8000 2. Check the status indicators (should all be green ✅) 3. Type your question in the input box 4. Click "🚀 Send Query" or press Enter 5. View the formatted response below 6. Use "📝 History" to see past queries --- ## 💻 **Interactive Command Line Guide** ### **Starting Interactive Mode:** ```bash python user_friendly_interface.py ``` ### **Available Commands:** - **help** - Show detailed help guide - **status** - Check system health - **history** - View query history - **clear** - Clear screen - **quit/exit** - Exit the interface ### **Example Session:** ``` 🎯 Your Query: Calculate 25 * 4 ⏳ Processing your query... ============================================================ 📤 QUERY: Calculate 25 * 4 ============================================================ 🤖 AGENT: math_agent ✅ STATUS: SUCCESS 🔢 ANSWER: 100.0 💾 MONGODB STORED: ❌ No 🕐 TIME: 12:08:31 ============================================================ ``` --- ## ⚡ **Quick Query Tool Guide** ### **Single Query Syntax:** ```bash python quick_query.py "Your question here" ``` ### **Examples:** ```bash # Math calculation python quick_query.py "Calculate 25 * 4" # Weather query python quick_query.py "What is the weather in Mumbai?" # Document analysis python quick_query.py "Analyze this text: Hello world" ``` ### **Output Format:** ``` 🚀 MCP QUICK QUERY ================================================== 📤 Query: Calculate 100 + 200 ================================================== ✅ Server: Ready ✅ MongoDB: Connected ✅ Agents: 3 loaded ⏳ Processing... 📊 RESULT: ------------------------------ 🤖 Agent: math_agent ✅ Status: SUCCESS 🔢 Answer: 300.0 💾 MongoDB: ❌ Not Stored 🕐 Time: 12:11:01 ✅ Query completed successfully! ``` --- ## 🤖 **Available Agents** ### **🔢 Math Agent** - **Triggers**: calculate, compute, math, +, -, *, /, % - **Capabilities**: Basic arithmetic, percentages, formulas - **Examples**: "Calculate 25 * 4", "What is 20% of 500?" ### **🌤️ Weather Agent** - **Triggers**: weather, temperature, forecast, climate - **Capabilities**: Real-time weather data, forecasts - **Examples**: "Weather in Mumbai", "Temperature in Delhi" ### **📄 Document Agent** - **Triggers**: analyze, document, text, process - **Capabilities**: Text analysis, content processing - **Examples**: "Analyze this text: Hello world" --- ## 💾 **MongoDB Integration** ### **What Gets Stored:** - All queries and responses - Agent processing results - Timestamps and metadata - Enhanced analytics data ### **Storage Features:** - **Real-time storage**: Every interaction saved - **Query history**: Track all past queries - **Agent analytics**: Performance metrics - **Enhanced functions**: Advanced storage capabilities ### **Access Stored Data:** ```python # Using enhanced storage functions from enhanced_mongodb_storage import get_agent_history, get_all_agent_stats # Get agent history history = get_agent_history("math_agent", limit=10) # Get statistics stats = get_all_agent_stats() ``` --- ## 🔧 **Troubleshooting** ### **Server Not Running:** ```bash # Start the server python production_mcp_server.py # Check if running curl http://localhost:8000/api/health ``` ### **MongoDB Issues:** - Check your .env file for correct MongoDB credentials - Verify internet connection for cloud MongoDB - Run: `python connect_agents_mongodb_fixed.py` ### **Agent Not Responding:** - Check agent status: http://localhost:8000/api/agents - Restart server: Stop and run `python production_mcp_server.py` - Check logs for error messages --- ## 📊 **System Monitoring** ### **Health Check:** - Web: http://localhost:8000/api/health - Command: `curl http://localhost:8000/api/health` ### **Agent Status:** - Web: http://localhost:8000/api/agents - Interactive: Type `status` in interactive mode ### **API Documentation:** - Full API docs: http://localhost:8000/docs --- ## 💡 **Tips for Best Results** ### **Query Writing:** - Be specific and clear - Use natural language - Include context when needed - Try different phrasings if needed ### **Math Queries:** - Use standard operators: +, -, *, /, % - Be explicit: "Calculate" or "What is" - Include units when relevant ### **Weather Queries:** - Use city names clearly - Try variations: "weather in", "temperature of" - Include country for ambiguous cities ### **Document Analysis:** - Prefix with "Analyze this text:" - Provide clear, readable content - Specify what type of analysis you want --- ## 🎉 **You're All Set!** Your user-friendly MCP system is ready to use with: - ✅ Multiple interfaces (web, interactive, quick query) - ✅ MongoDB storage and analytics - ✅ 3 intelligent agents (math, weather, document) - ✅ Real-time processing and responses - ✅ Query history and monitoring **Start exploring with any interface and enjoy your intelligent agent system!**