ai-agent-integration-with-mcp-tools

A smart financial assistant that combines Google Agent Development Kit with Model Context Protocol for real-time stock market data

🚀 Quick Start

# Clone and setup
git clone https://github.com/satendrakumar/ai-agent-integration-with-mcp-tools.git
cd ai-agent-integration-with-mcp-tools

# Install dependencies
uv sync

# Start MCP server
uv run -m mcp_server.server

# Configure environment (copy .env-example to .env)
cp .env-example .env

# Launch agent interface
adk web

Visit http://127.0.0.1:8000 and ask: "What's Apple's stock price today?"

💡 What This Does

Transform natural language into real-time financial insights:

• "Get me Tesla's current price" → Live TSLA data with daily changes
• "How is AAPL performing?" → Comprehensive Apple stock analysis
• "Show me Microsoft and Google prices" → Multi-stock comparison

🏗️ Architecture

┌─────────────────┐    HTTP/MCP     ┌──────────────────┐    API Call    ┌─────────────┐
│  Google ADK     │ ◄──────────────► │   MCP Server     │ ◄─────────────► │ Yahoo       │
│  Agent          │                 │  (Tools Layer)   │                │ Finance     │
│                 │                 │                  │                │             │
│ • NL Processing │                 │ • get_stock_price│                │ • Real-time │
│ • Conversation  │                 │ • Tool Discovery │                │ • Historical│
│ • Context Mgmt  │                 │ • Error Handling │                │ • Metadata  │
└─────────────────┘                 └──────────────────┘                └─────────────┘

Components

Component	Purpose	Technology
Agent Layer	Natural language understanding & conversation	Google ADK
Protocol Layer	Tool discovery & standardized communication	Model Context Protocol
Tools Layer	Stock data retrieval & processing	Python + yfinance
Data Layer	Real-time financial market data	Yahoo Finance API

📦 Installation

Prerequisites

• Python 3.11+
• UV package manager
• Ollama (for local LLM) or API access to hosted models

Step-by-Step Setup

1. Install Dependencies

   uv sync

2. Configure Environment

   cp .env-example .env

   Edit .env with your settings:

   AGENT_MODEL_NAME=ollama_chat/qwen3:1.7b
   API_BASE=http://localhost:11434
   MCP_SERVER_URL=http://localhost:8080/mcp

3. Start MCP Server

   uv run -m mcp_server.server

   Server runs on http://localhost:8080/mcp

4. Launch Agent Interface

   adk web

   Opens http://127.0.0.1:8000 in your browser

Verification

Test the MCP server independently:

npx @modelcontextprotocol/inspector

🛠️ Usage Examples

Basic Queries

User: "What's the current price of Apple?"
Agent: Apple (AAPL) is trading at $175.43, up $2.15 (+1.24%) today.

User: "Get Tesla's stock info"
Agent: Tesla (TSLA) is at $248.87, down $5.32 (-2.09%) in current session.

User: "How is MSFT performing?"
Agent: Microsoft (MSFT) shares are $378.91, up $4.55 (+1.22%) today.