🧠 Stock Analysis using MCP Local LLM (Ollama Qwen3) with LangChain

This project demonstrates how to build a fully local AI assistant that provides detailed stock analysis using:

• MCP (Model Context Protocol): Enables structured tool usage by the language model.

• Ollama: A tool for running large language models locally.(Qwen3)

• LangChain: A framework for developing applications powered by language models and run as MCP Client

• BeautifulSoup: For web scraping financial data from Screener.in.

📦 Features

🔍 Company Details: Retrieve company name, current price, market cap, PE ratio, ROE, ROCE, and more.

📈 Profit Analysis: Extract quarterly and yearly net profit data.

👥 Shareholding Patterns: Analyze holdings by promoters, DIIs, FIIs, and the public.

🔧 Tool Integration: Seamless integration with MCP tools for enhanced functionality.

⚙️ Configuration

MCP Server Setup

The MCP server is defined in mcp_config.json.

{
  "mcpServers": {
    "stock": {
      "command": "python",
      "args": ["StockMcp.py"],
      "transport": "stdio"
    }
  }
}


Give me the company details of CREDITACC.NS

🛠️ Project Structure

├── StockMcp.py         # MCP server with tool definitions
├── requirements.txt    # Python dependencies
├── README.md           # Project documentation

📚 Resources

MCP Github

Ollama Documentation

LangChain MCP Documentation

BeautifulSoup Documentation