Silver Price MCP Server

A Model Context Protocol (MCP) server that fetches live MCX India silver prices — plus a chart image — straight from Groww.in. No API key required; the data is scraped in real time and handed to any MCP-compatible client as structured JSON and a PNG chart.

What it does

• Live silver price — current price, open, previous close, and % / ₹ change, pulled straight from MCX India via Groww.in.

• 52-week range — high/low over the trailing year.

• Volume & open interest — lot size and OI lots included.

• Trend detection — flags the move as UP or DOWN.

• Chart image included — a full-page screenshot of the price chart, returned as a base64-encoded PNG alongside the JSON.

• No API key needed — works out of the box; the data comes from a headless Chrome scrape, not a paid feed.

Related MCP server: India Market MCP

Repository layout

├── src/                  # MCP server implementation (tool + scraper)
├── requirements.txt      # Python dependencies
├── pyproject.toml        # Project configuration
└── README.md             # This file

Requirements

• Python 3.10+

• Chrome browser (runs headless via Selenium)

• Dependencies: mcp, selenium, webdriver-manager

Quickstart

1. Install

git clone https://github.com/bhavuk1409/silver-mcx-mcp-server.git
cd silver-mcx-mcp-server

pip install -r requirements.txt

2. Test it directly

python test.py

3. Connect to Claude Desktop

Add to your Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "silver-price": {
      "command": "python",
      "args": ["-m", "mcp_metal_price"],
      "cwd": "/path/to/silver-mcx-mcp-server"
    }
  }
}

Restart Claude Desktop and ask: "Get the current silver price."

Tool

Example response:

{
  "metal": "Silver",
  "symbol": "XAG",
  "currency": "INR",
  "current_price": "2,42,798.00",
  "price_formatted": "₹2,42,798.00",
  "open": "2,42,775.00",
  "previous_close": "2,40,277.00",
  "52w_low": "1,09,741.00",
  "52w_high": "2,59,692.00",
  "oi_lots": "12,394",
  "lot_size": "30",
  "change": "+2521.00",
  "percent_change": "+1.05%",
  "trend": "UP"
}

How it works

1. Opens the Groww.in silver page in a headless Chrome instance via Selenium.

2. Waits ~8 seconds for the page content to fully load.

3. Extracts price data from the page using regex.

4. Captures a full-page screenshot of the chart.

5. Returns the parsed JSON plus the chart image to the MCP client.

Notes

• Scraping takes roughly 8–10 seconds per call.

• Chrome runs headless — no visible window.

• Chart image is PNG format, base64-encoded.

• Data source is MCX India via Groww.in and can change format if the site changes its layout.

License

MIT