mcp-yfinance logo

mcp-yfinance

by nicholasf

A simple wrapper around the yFinance library that operates as an MCP tool. It fetches stock data and passes back a Pandas Dataframe to the MCP host.

View on GitHub

Last updated: N/A

What is mcp-yfinance?

This is an MCP tool that wraps the yFinance library to fetch stock market data. It's designed to be used as an extension within an MCP host environment like Goose, allowing you to retrieve and analyze stock information using natural language queries through an integrated LLM.

How to use mcp-yfinance?

  1. Register the tool as an extension in your MCP host (e.g., Goose). 2. Configure the extension with a name, command to run (e.g., uv run server.py), and timeout. 3. Invoke the extension within a session using the configured name (e.g., goose session -n yf). 4. Use the extension's command (e.g., yf__history AAPL) to fetch stock data.

Key features of mcp-yfinance

  • Fetches stock data from yFinance

  • Operates as an MCP extension

  • Returns data as a Pandas DataFrame

  • Integrates with LLMs for natural language queries

  • Easy to install and configure

Use cases of mcp-yfinance

  • Fetching daily stock prices for individual instruments

  • Analyzing stock performance over specific time periods

  • Integrating stock data into LLM-powered applications

  • Demonstrating MCP extension development

  • Proof of concept for data retrieval within an MCP environment

FAQ from mcp-yfinance

What is yFinance?

yFinance is a popular Python library for retrieving stock market data from Yahoo Finance.

What is an MCP?

MCP stands for Meta-Cognitive Programming. It's a framework for building intelligent agents that can reason and interact with the world.

What is Goose?

Goose is an MCP host that allows you to run and manage MCP extensions.

What kind of data does this tool provide?

The tool provides historical stock data, including Open, High, Low, Close prices, Volume, Dividends, and Stock Splits.

What are the limitations of this tool?

The tool's performance may be limited by token limits of the integrated LLM when performing complex aggregations.