A MCP server (Model Context Protocol) for accessing Indian stock market data and integrating it with Claude Desktop/Any other AI.
#mcp-server-indian-stock-market
An MCP (Model Context Protocol) server for accessing Indian stock market data and integrating it with Claude via the MCP framework.
=================================================================================
📋 Requirements
To use this project, make sure you have the following:
A Claude.ai account (MCP support is available for all account types)
Claude Desktop App (macOS or Windows)
A code editor such as Visual Studio Code
uv – a fast, Rust-based Python package manager
Install uv
macOS (Homebrew):
brew install uv
Windows (WinGet):
winget install --id=astral-sh.uv -e
=================================================================================
🚀 Project Setup
Follow the steps below to set up the MCP server locally.
-
Create Project Directory mkdir mcp-server-indian-stock-market cd mcp-server-indian-stock-market
-
Initialize a uv Project uv init
-
Create a Virtual Environment uv venv
-
Activate the Virtual Environment
macOS / Linux: source .venv/bin/activate
Windows: Command Prompt .venv\Scripts\activate.bat
PowerShell .venv\Scripts\Activate.ps1
Git Bash source .venv/Scripts/activate
To deactivate the environment: deactivate
=================================================================================
📦 Install Dependencies Install the MCP Python SDK with CLI support and required dependencies: uv add "mcp[cli]" httpx
=================================================================================
📁 Add main.py
Clone the repository or download the source code: git clone
Copy main.py into your project root directory: cp /main.py .
Your directory structure should look like this:
mcp-server-indian-stock-market/ ├── main.py ├── pyproject.toml └── .venv/
=================================================================================
⚙️ MCP Configuration
Add the following entry to your MCP configuration file:
{ "indian_stock_market": { "command": "/Users/username/.local/bin/uv", "args": [ "run", "--with", "mcp[cli]", "mcp", "run", "/main.py" ] } }
Configuration Notes Replace username with your system username Replace /main.py with the full path to main.py Ensure the virtual environment is active when running the server uv must exist at the specified path
✅ You're Ready Once configured, restart Claude Desktop and the MCP server will be available for use.
=================================================================================
🛠️ Optional Improvements
You may want to add: Tool descriptions exposed by the MCP server Example Claude prompts Error handling and logging API rate-limit handling