US Stock Data MCP Server
@mingyaw
Overview
What is US Stock Data MCP Server?
The US Stock Data MCP Server is a Model Context Protocol (MCP) server designed for accessing and updating historical price data for US stocks. It stores data locally in CSV format and automatically retrieves updates from Yahoo Finance. This server is intended for developers and AI agents that need local access to US stock market data.
How to use US Stock Data MCP Server?
Install Python 3.x, clone the repository, and run pip install -r requirements.txt. Start the server with python server.py. The default data storage path is ~/Library/Application Support/us-market-data/data, which can be customized using the US_STOCK_DATA_DIR environment variable. Use the MCP tools get_local_stock_data and update_stock_data to retrieve and update historical data.
Key features of US Stock Data MCP Server
- Local CSV storage for quick access
- Automatic data updates from Yahoo Finance
- Safe atomic writes using temporary files
- Customizable start date for updates
- Duplicate data automatically handled
- 5-second delay between updates to avoid rate limits
Use cases of US Stock Data MCP Server
- Fetch and store historical stock data for offline analysis
- Automatically update daily stock price data for a portfolio tracker
- Provide historical data to an AI agent for market research
- Replace external API calls with local data for backtesting strategies
FAQ from US Stock Data MCP Server
What data source does this server use?
The server retrieves stock data from Yahoo Finance via the yfinance Python library.
Where is the stock data stored locally?
Data is stored in CSV format at ~/Library/Application Support/us-market-data/data by default. You can change this with the US_STOCK_DATA_DIR environment variable.
How do I update data for a specific stock?
Use the update_stock_data tool with the stock symbol (e.g., AAPL) and an optional start date in YYYY-MM-DD format. The default start date is 2015-01-01.
Does the server handle duplicate records?
Yes, duplicate data is automatically handled during updates, keeping only the latest records.
Is there a rate limit for updates?
A 5-second delay is enforced between update operations to avoid frequent API requests to Yahoo Finance.
