The Finance MCP Server delivers a comprehensive suite of stock market data and technical analysis tools that can be queried via the MCP protocol. It exposes capabilities such as getting comprehensive stock information, historical data across multiple timeframes, real-time quotes, options data, institutional holders, earnings calendars, analyst recommendations, financial statements, news, technical analysis indicators, sector performance, and dividend history. These tools enable Claude Desktop and other MCP-compatible clients to request detailed financial insights, perform technical analysis, and compare multiple securities with caching to improve response times. You can call tools like get_comprehensive_stock_info, get_historical_data, get_technical_analysis, and get_dividend_history by sending MCP JSON payloads that specify the tool name and the necessary arguments (e.g., a ticker symbol).

To use the server from Claude Desktop or an MCP client, ensure the Finance MCP Server is reachable and configured in Claude’s MCP settings. A typical tool call looks like { "tool": "get_comprehensive_stock_info", "arguments": {"symbol": "AAPL"} }. You can also fetch historical data with { "tool": "get_historical_data", "arguments": {"symbol": "AAPL", "period": "1y", "interval": "1d"} } and retrieve technical indicators with { "tool": "get_technical_analysis", "arguments": {"symbol": "AAPL", "indicators": ["RSI", "MACD", "SMA"]} }.

Prerequisites:

• Python 3.8 or higher
• Git
• Pip (typically bundled with Python)

Step-by-step installation:

1. Clone the repository: git clone https://github.com/akshatbindal/finance-mcp-server.git cd finance-mcp-server

2. Create and activate a virtual environment (recommended): python -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate

3. Install in development mode (for local development): pip install -e .

4. Install necessary development dependencies (optional for testing): pip install pytest black flake8 mypy

5. Start the server (example): python -m finance_mcp.server

   or use the console script if installed:

   finance-mcp-server

6. Optional: run with debug logging: LOG_LEVEL=DEBUG python -m finance_mcp.server

7. Configuration in Claude Desktop or MC P clients: Use the provided mcpServers configuration (see README) and point to the command that runs the server (in this setup, the Python module invocation).

Tips and common considerations:

• Python 3.8+ is required; ensure your environment matches this.
• The server caching is configurable; use env variables like LOG_LEVEL and CACHE_DURATION to tune performance.
• If you run into issues starting the server, verify that your Python environment has the required dependencies (mcp >= 1.0.0, yfinance, pandas, ta) installed.
• For Claude Desktop integration, ensure your JSON configuration is valid and the server command path is correct. Restart Claude Desktop after changes.
• The server retrieves real-time data from Yahoo Finance and does not store user data locally, aligning with privacy and security best practices.