AMFI MCP Server

An open-source Model Context Protocol (MCP) server that gives any AI assistant access to official Indian mutual-fund NAV data from AMFI (the Association of Mutual Funds in India).

Ask Claude, GitHub Copilot, Cursor, or any MCP-compatible client things like "What's the latest NAV of Parag Parikh Flexi Cap?" and get a real, current answer — using AMFI's official, free, no-API-key feed.

Why this exists

AI models don't know today's mutual-fund NAVs, and most existing market-data MCP servers cover US stocks — not Indian mutual funds. This server fills that gap with the authoritative Indian source, and needs no API key.

Tools

Tool	What it does
search_mutual_funds(query)	Find schemes by name (e.g. "hdfc small cap"), returning each scheme's code and latest NAV.
get_fund_nav(scheme_code)	Get the latest official NAV and date for one scheme, by its numeric AMFI code.

Data source: AMFI's daily NAVAll feed.

Install

git clone https://github.com/SetuAI/amfi-mcp-server.git
cd amfi-mcp-server
pip install -e .

Run

amfi-mcp

The server communicates over stdio, so it will sit quietly waiting for an MCP client to connect. To inspect it by hand:

npx @modelcontextprotocol/inspector amfi-mcp

Then: Connect → Tools → List Tools → try search_mutual_funds with hdfc small cap, then get_fund_nav with a scheme code from the results.

Connect it to an AI client

Claude Desktop

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "amfi": {
      "command": "amfi-mcp"
    }
  }
}

(macOS: ~/Library/Application Support/Claude/claude_desktop_config.json, Windows: %APPDATA%\Claude\claude_desktop_config.json. Fully quit and reopen Claude Desktop afterwards.)

VS Code (GitHub Copilot)

Add to .vscode/mcp.json in your workspace:

{
  "servers": {
    "amfi": {
      "command": "amfi-mcp"
    }
  }
}

Click Start, open Copilot Chat, switch to Agent mode, and ask a mutual-fund question.

Example prompts

• "Search for HDFC small cap funds and give me their latest NAVs."
• "What's the NAV of scheme code 122639?"
• "Find Parag Parikh flexi cap and tell me today's value."

How it works

• amfi_client.py downloads and parses AMFI's NAV file, caches it in memory (refreshing at most once per hour), and offers the lookups.
• server.py wraps those lookups as MCP tools with FastMCP.

The NAV file updates once per business day, so values are accurate as of AMFI's last publishing day. This project is not affiliated with AMFI; it simply reads AMFI's public feed. Data is provided as-is, for informational purposes, and is not investment advice.

Requirements

• Python 3.10+
• Outbound internet access to www.amfiindia.com

License

MIT — see LICENSE.