The "sagar_builds" Groww MCP Server Guide

If you've ever wished you could just ask Claude about your Groww portfolio instead of navigating dashboards, this guide is for you. The Groww Python MCP Server bridges your Groww trading account directly into Claude — letting you query positions, prices, and data using plain English.

This guide walks you through the full setup: cloning the repo, configuring your API keys, starting the server, and wiring it up to Claude Desktop.

📺 Video Guide

Watch on YouTube if the player doesn't load

🚀 What Is an MCP Server?

MCP (Model Context Protocol) is an open standard that lets AI models like Claude connect to external tools and data sources. By running your own Groww MCP server locally, Claude can read your Groww data in real time — all from within the chat interface, no copy-pasting required.

🛠️ Step-by-Step Setup

Step 1: Get the Code

If you have Git installed, clone the repository:

git clone https://github.com/SAGAR-TAMANG/groww-python-mcp
cd groww-python-mcp

No Git? Head to https://github.com/SAGAR-TAMANG/groww-python-mcp, click the green Code button, and select Download ZIP. Unzip it and open the folder.

Step 2: Open in Your Terminal or VS Code

Open a terminal inside the project folder, or open the folder in VS Code and use its integrated terminal (Ctrl + `` ``).

Step 3: Install `uv` (Python Package Manager)

This project uses uv — a fast, modern Python package and project manager. If you don't have it yet, install it from the official docs:

👉 https://docs.astral.sh/uv/getting-started/installation/

Follow the instructions for your operating system (Windows, macOS, or Linux). Once installed, verify it works:

uv --version

Step 4: Get Your Groww API Keys

Log in to Groww and navigate to the Trade API section:
👉 https://groww.in/trade-api/api-keys
Sign up for the free plan — it's available for the first few months at no cost.
Generate your API Key and API Secret Key. Keep these safe — you'll need them in the next step.

Step 5: Configure Your `.env` File

In the project root, you'll find a .env file (or a .env.example to copy from). Open it and fill in your credentials:

GROWW_API_KEY=your_api_key_here
GROWW_API_SECRET=your_api_secret_here

Save the file. These keys stay local on your machine — they are never sent anywhere except to Groww's own API.

Step 6: Start the MCP Server

Run the server with a single command:

uv run main.py

You should see the server start up on port 8000. The server will be live at:

http://127.0.0.1:8000

Keep this terminal window open while you use Claude. The server needs to be running for Claude to access it.

Step 7: Connect Claude Desktop to the Server

Now tell Claude Desktop about your new server. Open Claude's developer settings and locate the config file — it's typically called claude_desktop_config.json.

Add the following entry inside the mcpServers object:

"mcpServers": {
  "groww_mcp": {
    "command": "npx",
    "args": [
      "-y",
      "mcp-remote",
      "http://127.0.0.1:8000/sse"
    ]
  }
}

Note: If mcpServers already exists in your config, just add the "groww_mcp" block inside it — don't duplicate the outer key.

Restart Claude Desktop. You should now see Groww available as a connected tool inside Claude.

✅ You're Connected!

Once restarted, Claude can use the Groww MCP server to answer questions about your account. Try prompts like:

"What does my current portfolio look like?"
"What's the latest price of INFY?"
"Show me my open positions."

🔌 Using With Other MCP Clients

The server isn't exclusive to Claude Desktop. Since it's a standard SSE-based MCP server running on http://127.0.0.1:8000/sse, you can connect it from any MCP-compatible client — other AI tools, custom scripts, or automation workflows. Just point them at the same SSE endpoint.

🤝 Contributing

This is an open-source project. If you'd like to add features — like natural language order placement, sector breakdowns, or P&L summaries — open a PR on GitHub!

Built by Sagar Tamang

Connect Claude to Your Groww Account: The Python MCP Server Guide