🤖 AI-MCP Orchestrator

Next-Generation Agentic AI System with 17+ MCP Tools

A production-ready ChatGPT-style system integrating LangChain, LangGraph, FastMCP, and Google Gemini 2.0 Flash with 17+ powerful MCP tools, intelligent routing, persistent memory, and a beautiful reactive UI.

Features • Architecture • Installation • Usage • API Docs

---

📋 Table of Contents

• Overview
• Key Features
• System Architecture
• MCP Tools
• Technology Stack
• Installation
• Configuration
• Usage Guide
• API Documentation
• Project Structure
• Development
• Troubleshooting
• Contributing
• License

---

🌟 Overview

AI-MCP Orchestrator is a sophisticated agentic AI system that combines the power of Google's Gemini 2.0 Flash model with an extensive suite of Model Context Protocol (MCP) tools. Built on LangChain and LangGraph, it provides intelligent tool routing, multi-step reasoning, and seamless integration with external services.

What Makes It Special?

• 🎯 17+ MCP Tools - From web search to system monitoring, weather to translation
• 🧠 Intelligent Routing - Automatically selects the right tools for each query
• 🔄 LangGraph Pipeline - Multi-agent orchestration with state management
• 💾 Persistent Memory - Conversation history and context retention
• 🎨 Beautiful UI - Modern, responsive interface with real-time updates
• ⚡ Real-time WebSocket - Streaming responses and live tool execution
• 🔧 Toggleable Tools - Enable/disable tools on-the-fly
• 📊 Analytics Dashboard - Monitor system performance and tool usage

---

✨ Features

🤖 AI Capabilities

• Multi-turn Conversations with context awareness
• Tool Chaining - Combine multiple tools in a single query
• Intelligent Fallbacks - Graceful degradation when tools fail
• Streaming Responses - Real-time token-by-token output
• Memory Management - Short-term and long-term context

🛠️ MCP Tools Suite

• Web Search - DuckDuckGo integration
• File Management - Read, write, organize files
• Database Operations - MongoDB CRUD operations
• Email Automation - Send and manage emails
• Google Drive - Cloud storage integration
• Workflow Automation - Schedule and execute tasks
• Knowledge Base - Store and retrieve information
• Analytics - Data analysis and visualization
• API Integration - Connect to external APIs
• Weather/Climate - Real-time weather data
• Wikipedia - Encyclopedia search and retrieval
• Python Execution - Safe code execution sandbox
• Screen Monitoring - Screenshots and display info
• System Monitoring - CPU, memory, disk metrics
• Calculator - Math operations and unit conversion
• Translator - Multi-language translation

🎨 User Interface

• ChatGPT-style Interface - Familiar and intuitive
• Tool Status Cards - Visual tool management
• Animated Particles - Dynamic background effects
• Dark Theme - Eye-friendly design
• Responsive Layout - Works on all devices
• Message History - Scroll through past conversations
• Typing Indicators - Real-time feedback

---

🏗️ Architecture

High-Level System Design

┌─────────────────────────────────────────────────────────────────┐
│                         FRONTEND (React)                         │
│  ┌──────────────┐  ┌──────────────┐  ┌──────────────┐         │
│  │   Chat UI    │  │  Tool Panel  │  │  Analytics   │         │
│  └──────┬───────┘  └──────┬───────┘  └──────┬───────┘         │
│         │                  │                  │                  │
│         └──────────────────┴──────────────────┘                  │
│                            │                                      │
│                    WebSocket / REST API                          │
└────────────────────────────┼────────────────────────────────────┘
                             │
┌────────────────────────────┼────────────────────────────────────┐
│                    BACKEND (FastAPI)                             │
│  ┌──────────────────────────┴──────────────────────────┐        │
│  │              API Layer (Routes)                      │        │
│  │  • Chat Routes    • MCP Routes    • Memory Routes   │        │
│  └──────────────────────┬──────────────────────────────┘        │
│                         │                                         │
│  ┌──────────────────────┴──────────────────────────────┐        │
│  │         LangGraph Orchestration Pipeline             │        │
│  │  ┌──────────┐  ┌──────────┐  ┌──────────┐          │        │
│  │  │  Router  │→ │ Executor │→ │ Aggregator│          │        │
│  │  └──────────┘  └──────────┘  └──────────┘          │        │
│  └──────────────────────┬──────────────────────────────┘        │
│                         │                                         │
│  ┌──────────────────────┴──────────────────────────────┐        │
│  │              Tool Registry (17 Tools)                │        │
│  │  ┌────────┐ ┌────────┐ ┌────────┐ ┌────────┐       │        │
│  │  │Web Srch│ │File Mgr│ │Database│ │ Email  │ ...   │        │
│  │  └────────┘ └────────┘ └────────┘ └────────┘       │        │
│  └──────────────────────┬──────────────────────────────┘        │
│                         │                                         │
│  ┌──────────────────────┴──────────────────────────────┐        │
│  │            Services & Integrations                   │        │
│  │  • Database Service (MongoDB)                        │        │
│  │  • Gemini AI Service (Google)                        │        │
│  │  • Memory Service (Context Management)               │        │
│  └──────────────────────────────────────────────────────┘        │
└──────────────────────────────────────────────────────────────────┘
                             │
┌────────────────────────────┼────────────────────────────────────┐
│                    EXTERNAL SERVICES                             │
│  ┌──────────┐  ┌──────────┐  ┌──────────┐  ┌──────────┐       │
│  │ Gemini   │  │ MongoDB  │  │  Google  │  │   APIs   │       │
│  │ 2.0 Flash│  │          │  │  Drive   │  │          │       │
│  └──────────┘  └──────────┘  └──────────┘  └──────────┘       │
└──────────────────────────────────────────────────────────────────┘

LangGraph Pipeline Flow

User Query
    ↓
┌───────────────┐
│  Query Router │ ← Analyzes query and selects appropriate tools
└───────┬───────┘
        ↓
┌───────────────┐
│ Tool Executor │ ← Executes selected MCP tools in parallel/sequence
└───────┬───────┘
        ↓
┌───────────────┐
│  Aggregator   │ ← Combines results from multiple tools
└───────┬───────┘
        ↓
┌───────────────┐
│ Response Gen  │ ← Generates natural language response
└───────┬───────┘
        ↓
   User Response

Data Flow Architecture

┌──────────────┐
│   User Input │
└──────┬───────┘
       ↓
┌──────────────────────────────────────┐
│  1. WebSocket Connection Established │
└──────┬───────────────────────────────┘
       ↓
┌──────────────────────────────────────┐
│  2. Query Preprocessing               │
│     • Tokenization                    │
│     • Context Injection               │
│     • Memory Retrieval                │
└──────┬───────────────────────────────┘
       ↓
┌──────────────────────────────────────┐
│  3. Tool Selection (Router)           │
│     • Keyword Matching                │
│     • Semantic Analysis               │
│     • Multi-tool Detection            │
└──────┬───────────────────────────────┘
       ↓
┌──────────────────────────────────────┐
│  4. Tool Execution                    │
│     • Parallel Execution              │
│     • Error Handling                  │
│     • Result Caching                  │
└──────┬───────────────────────────────┘
       ↓
┌──────────────────────────────────────┐
│  5. Response Generation               │
│     • Result Aggregation              │
│     • Gemini 2.0 Processing           │
│     • Streaming Output                │
└──────┬───────────────────────────────┘
       ↓
┌──────────────────────────────────────┐
│  6. Memory Update                     │
│     • Conversation Storage            │
│     • Context Update                  │
└──────┬───────────────────────────────┘
       ↓
┌──────────────┐
│ User Response│
└──────────────┘

---

🧰 MCP Tools (17)

1. Web Search 🔍

• Provider: DuckDuckGo
• Actions: Search web, get snippets, extract URLs
• Use Cases: Research, fact-checking, current events

2. File Manager 📁

• Actions: Read, write, delete, list files
• Supported: Text, JSON, CSV, Markdown
• Use Cases: Document management, data processing

3. Database 🗄️

• Provider: MongoDB
• Actions: CRUD operations, queries, aggregations
• Use Cases: Data storage, retrieval, analytics

4. Email 📧

• Protocol: SMTP
• Actions: Send emails, attachments, templates
• Use Cases: Notifications, reports, communication

5. Google Drive ☁️

• Actions: Upload, download, share, organize
• Use Cases: Cloud storage, collaboration

6. Automation ⚡

• Actions: Schedule tasks, workflows, triggers
• Use Cases: Recurring tasks, batch processing

7. Memory 🧠

• Actions: Store, recall, search context
• Use Cases: Conversation history, user preferences

8. Analytics 📊

• Actions: Data analysis, visualization, trends
• Use Cases: Business intelligence, reporting

9. Knowledge Base 📚

• Actions: Store articles, search, categorize
• Use Cases: Documentation, wiki, FAQs

10. API Integration 🔗

• Actions: REST calls, webhooks, OAuth
• Use Cases: Third-party integrations

11. Climate/Weather 🌤️

• Provider: Python Weather API
• Actions: Current weather, forecasts, climate data
• Use Cases: Travel planning, weather alerts

12. Wikipedia 📖

• Actions: Search, summaries, full articles
• Use Cases: Research, definitions, learning

13. Python Code 💻

• Actions: Execute code, validate syntax, analyze
• Use Cases: Calculations, data processing, automation

14. Screen Monitor 🖥️

• Actions: Screenshots, display info, resolution
• Use Cases: Documentation, troubleshooting

15. System Monitor 📈

• Actions: CPU, memory, disk, network stats
• Use Cases: Performance monitoring, diagnostics

16. Calculator 🔢

• Actions: Math operations, unit conversions
• Use Cases: Calculations, conversions, equations

17. Translator 🌐

• Actions: Translate text, detect language
• Supported: 12+ languages
• Use Cases: Multilingual communication

---

🔧 Technology Stack

Backend

• Framework: FastAPI 0.120.0
• AI/ML:
  • Google Gemini 2.0 Flash
  • LangChain (latest)
  • LangGraph (latest)
• Database: MongoDB (Motor async driver)
• WebSocket: FastAPI WebSockets
• Validation: Pydantic v2

Frontend

• Framework: React 18
• Build Tool: Vite
• Styling: TailwindCSS
• Animations: Framer Motion
• Icons: Lucide React
• HTTP Client: Axios

DevOps

• Environment: Python 3.13+
• Package Manager: pip, npm
• Process Manager: Uvicorn
• Logging: Python logging with colors

---

📦 Installation

Prerequisites

• Python 3.13+
• Node.js 18+
• MongoDB (local or cloud)
• Git

Step 1: Clone Repository

git clone https://github.com/yourusername/ai-mcp-orchestrator.git
cd ai-mcp-orchestrator

Step 2: Backend Setup

cd backend

# Create virtual environment
python -m venv venv

# Activate virtual environment
# Windows:
venv\Scripts\activate
# Linux/Mac:
source venv/bin/activate

# Install dependencies
pip install -r requirements.txt

Step 3: Frontend Setup

cd frontend

# Install dependencies
npm install

Step 4: Environment Configuration

Create .env file in backend/ directory:

# API Keys
GEMINI_API_KEY=your_gemini_api_key_here
OPENAI_API_KEY=

# Model Configuration
GEMINI_MODEL=gemini-2.0-flash-exp
TEMPERATURE=0.7
MAX_TOKENS=8192

# Database
MONGODB_URL=mongodb://localhost:27017
DATABASE_NAME=ai_mcp_orchestrator

# Server
HOST=0.0.0.0
PORT=8000
DEBUG=True

# CORS (comma-separated)
CORS_ORIGINS=http://localhost:3000,http://localhost:5173

# MCP Configuration
MCP_TIMEOUT=30
MAX_CONCURRENT_TOOLS=5

Step 5: Start MongoDB

# Windows (if installed as service)
net start MongoDB

# Linux/Mac
sudo systemctl start mongod

# Or use MongoDB Atlas (cloud)

---

🚀 Usage Guide

Starting the Application

Option 1: Manual Start

Terminal 1 - Backend:

cd backend
python main.py

Terminal 2 - Frontend:

cd frontend
npm run dev

Option 2: Using Start Scripts

Windows:

start.bat

Linux/Mac:

chmod +x start.sh
./start.sh

Accessing the Application

• Frontend: http://localhost:3000
• Backend API: http://localhost:8000
• API Docs: http://localhost:8000/docs
• Health Check: http://localhost:8000/health

Using the Chat Interface

1. Open http://localhost:3000
2. Enable Tools - Click "MCP Tools" button and toggle desired tools
3. Start Chatting - Type your query in the input box
4. View Results - See AI responses with tool execution details

Example Queries

🔍 Web Search:
"Search for latest AI news"

🌤️ Weather:
"What's the weather in Tokyo?"

📖 Wikipedia:
"Tell me about quantum computing"

💻 Python Code:
"Run this code: print('Hello World')"

📊 System Monitor:
"Show me my CPU usage"

🔢 Calculator:
"Calculate 25 * 4 + 10"

🌐 Translator:
"Translate 'Good morning' to Spanish"

🔗 Multi-Tool:
"Search for Python tutorials and save the results to a file"

---

📚 API Documentation

REST Endpoints

Chat API

POST /api/chat/message

{
  "message": "Your query here",
  "session_id": "optional-session-id",
  "active_tools": ["web_search", "calculator"]
}

Response:

{
  "response": "AI generated response",
  "tools_used": ["web_search"],
  "execution_time": 1.23,
  "session_id": "session-123"
}

MCP Tools API

GET /api/mcp/tools

{
  "tools": [
    {
      "name": "web_search",
      "description": "Search the web",
      "enabled": true
    }
  ]
}

POST /api/mcp/tools/{tool_name}/toggle

{
  "enabled": true
}

Memory API

GET /api/memory/history?session_id=123&limit=10

DELETE /api/memory/clear?session_id=123

WebSocket API

Connect: ws://localhost:8000/api/ws/chat

Send Message:

{
  "type": "message",
  "content": "Your query",
  "session_id": "session-123"
}

Receive Response:

{
  "type": "response",
  "content": "AI response",
  "tools_used": ["web_search"],
  "done": true
}

---

📁 Project Structure

ai-mcp-orchestrator/
├── backend/
│   ├── config.py              # Configuration settings
│   ├── main.py                # FastAPI application entry
│   ├── requirements.txt       # Python dependencies
│   │
│   ├── routes/                # API route handlers
│   │   ├── chat_routes.py
│   │   ├── mcp_routes.py
│   │   ├── memory_routes.py
│   │   └── websocket_routes.py
│   │
│   ├── services/              # Business logic services
│   │   ├── db_service.py
│   │   ├── gemini_service.py
│   │   └── memory_service.py
│   │
│   ├── mcp_tools/             # 17 MCP tool implementations
│   │   ├── web_search_mcp.py
│   │   ├── file_manager_mcp.py
│   │   ├── climate_mcp.py
│   │   ├── wikipedia_mcp.py
│   │   └── ... (13 more)
│   │
│   ├── langgraph_pipeline/    # LangGraph orchestration
│   │   ├── graph_builder.py
│   │   ├── router.py
│   │   ├── executor.py
│   │   └── aggregator.py
│   │
│   └── utils/                 # Utility functions
│       ├── tool_registry.py
│       └── helpers.py
│
├── frontend/
│   ├── package.json
│   ├── vite.config.js
│   ├── tailwind.config.js
│   │
│   ├── src/
│   │   ├── App.jsx            # Main application
│   │   ├── main.jsx           # Entry point
│   │   │
│   │   ├── components/        # React components
│   │   │   ├── ChatInterface.jsx
│   │   │   ├── MCPToolPanel.jsx
│   │   │   ├── MCPCard.jsx
│   │   │   └── ParticleBackground.jsx
│   │   │
│   │   ├── context/           # React context
│   │   │   └── MCPContext.jsx
│   │   │
│   │   ├── services/          # API services
│   │   │   └── api.js
│   │   │
│   │   └── styles/            # CSS styles
│   │       └── index.css
│   │
│   └── public/                # Static assets
│
├── README.md                  # This file
├── UPDATE_SUMMARY.md          # Recent updates
└── .gitignore

---

🛠️ Development

Adding a New MCP Tool

1. Create tool file: backend/mcp_tools/your_tool_mcp.py

class YourTool:
    def __init__(self):
        self.name = "your_tool"
        self.description = "Tool description"
        self.enabled = True
    
    async def execute(self, action: str, **kwargs):
        # Implementation
        return {"success": True, "data": "result"}
    
    def get_schema(self):
        return {
            "name": self.name,
            "description": self.description,
            "parameters": {...}
        }

1. Register in __init__.py:

from .your_tool_mcp import YourTool
__all__ = [..., "YourTool"]

1. Add to tool registry: utils/tool_registry.py

2. Add keywords: langgraph_pipeline/router.py

3. Add frontend icon: frontend/src/components/MCPCard.jsx

Running Tests

# Backend tests
cd backend
pytest

# Frontend tests
cd frontend
npm test

Code Formatting

# Backend
black backend/
flake8 backend/

# Frontend
npm run lint
npm run format

---

🐛 Troubleshooting

Common Issues

1. CORS Error

Error: CORS policy blocked

Solution: Check CORS_ORIGINS in .env matches your frontend URL

2. MongoDB Connection Failed

Error: Could not connect to MongoDB

Solution:

• Ensure MongoDB is running
• Check MONGODB_URL in .env
• Try: mongodb://localhost:27017 or MongoDB Atlas URL

3. Gemini API Error

Error: Invalid API key

Solution:

• Get API key from https://ai.google.dev
• Update GEMINI_API_KEY in .env

4. Module Not Found

ModuleNotFoundError: No module named 'fastapi'

Solution:

cd backend
pip install -r requirements.txt

5. Port Already in Use

Error: Address already in use

Solution:

• Change PORT in .env
• Or kill process: lsof -ti:8000 | xargs kill -9 (Mac/Linux)

Logs Location

• Backend: Console output with colored logging
• Frontend: Browser console (F12)
• MongoDB: Check MongoDB logs

---

🤝 Contributing

We welcome contributions! Please follow these steps:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to branch (git push origin feature/amazing-feature)
5. Open a Pull Request

Contribution Guidelines

• Follow existing code style
• Add tests for new features
• Update documentation
• Keep commits atomic and well-described

---

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

---

🙏 Acknowledgments

• Google - Gemini 2.0 Flash API
• LangChain - AI orchestration framework
• FastAPI - Modern Python web framework
• React - UI library
• MongoDB - Database solution

---

📞 Support

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Email: your.email@example.com

---

🗺️ Roadmap

• Add more MCP tools (20+ target)
• Voice input/output support
• Multi-user authentication
• Tool marketplace
• Mobile app (React Native)
• Docker containerization
• Kubernetes deployment
• Advanced analytics dashboard
• Plugin system for custom tools
• RAG (Retrieval Augmented Generation)

---

Built with ❤️ by the AI-MCP Team

⭐ Star us on GitHub if you find this useful!