investor-agent: A Financial Analysis MCP Server

Overview

The investor-agent is a Model Context Protocol (MCP) server that provides comprehensive financial insights and analysis to Large Language Models. It leverages real-time market data, news, and advanced analytics to help users obtain:

• Detailed ticker reports including company overview, news, key metrics, performance, dates, analyst recommendations, and upgrades/downgrades.
• Options data highlighting high open interest.
• Historical price trends for stocks.
• Essential financial statements (income, balance sheet, cash flow) formatted in millions USD.
• Up-to-date institutional ownership and mutual fund holdings.
• Earnings history and insider trading activity.

The server integrates with yfinance for market data retrieval.

Combine this with an MCP server for placing trades on a brokerage platform such as tasty-agent to place trades on tastytrade platform. Make sure to also enable web search functionality if you would like to incoporate latest news in your analysis.

Prerequisites

• Python: 3.12 or higher
• Package Manager: uv

Installation

First, install uv if you haven't already:

curl -LsSf https://astral.sh/uv/install.sh | sh

Then, you can run the investor-agent MCP server using uvx:

uvx investor-agent

Tools

The investor-agent server comes with several tools to support financial analysis:

Ticker Information

1. get_ticker_data

• Description: Retrieves a comprehensive report for a given ticker symbol, including company overview, news, key metrics, performance, dates, analyst recommendations, and upgrades/downgrades.
• Input:
  • ticker (string): Stock ticker symbol (e.g., "AAPL").
• Return: A formatted multi-section report.

1. get_available_options

• Description: Provides a list of stock options with the highest open interest.
• Inputs:
  • ticker_symbol (string): Stock ticker symbol.
  • num_options (int, optional): Number of options to return (default: 10).
  • start_date & end_date (string, optional): Date range in YYYY-MM-DD format.
  • strike_lower & strike_upper (float, optional): Desired strike price range.
  • option_type (string, optional): Option type ("C" for calls, "P" for puts).
• Return: A formatted table of options data.

1. get_price_history

• Description: Retrieves historical price data for a specific ticker.
• Inputs:
  • ticker (string): Stock ticker symbol.
  • period (string): Time period (choose from "1d", "5d", "1mo", "3mo", "6mo", "1y", "2y", "5y", "10y", "ytd", "max").
• Return: A table showing price history.

Financial Data Tools

1. get_financial_statements

• Description: Fetches financial statements (income, balance, or cash flow) formatted in millions USD.
• Inputs:
  • ticker (string): Stock ticker symbol.
  • statement_type (string): "income", "balance", or "cash".
  • frequency (string): "quarterly" or "annual".
• Return: A formatted financial statement.

1. get_institutional_holders

• Description: Retrieves details about major institutional and mutual fund holders.
• Input:
  • ticker (string): Stock ticker symbol.
• Return: Two formatted tables listing institutional and mutual fund holders.

1. get_earnings_history

• Description: Retrieves a formatted table of earnings history.
• Input:
  • ticker (string): Stock ticker symbol.
• Return: A table displaying historical earnings data.

1. get_insider_trades

• Description: Fetches the recent insider trading activity for a given ticker.
• Input:
  • ticker (string): Stock ticker symbol.
• Return: A formatted table showing insider trades.

Usage with MCP Clients

To integrate investor-agent with an MCP client (for example, Claude Desktop), add the following configuration to your claude_desktop_config.json:

{
  "mcpServers": {
    "investor": {
        "command": "path/to/uvx/command/uvx",
        "args": ["investor-agent"],
    }
  }
}

Debugging

You can leverage the MCP inspector to debug the server:

npx @modelcontextprotocol/inspector uvx investor-agent

For log monitoring, check the following directories:

• macOS: ~/Library/Logs/Claude/mcp*.log
• Windows: %APPDATA%\Claude\logs\mcp*.log

Development

For local development and testing:

1. Use the MCP inspector as described in the Debugging section.
2. Test using Claude Desktop with this configuration:

{
  "mcpServers": {
    "investor": {
      "command": "path/to/uv/command/uv",
      "args": ["--directory", "path/to/investor-agent", "run", "investor-agent"],
    }
  }
}

License

This MCP server is licensed under the MIT License. See the LICENSE file for details.