Financial Analyst MCP Server

A small MCP (Model Context Protocol) server that gives Claude Desktop, Cursor, or any other MCP-compatible client live tools for stock research:

• get_stock_quote — current price, daily change, day range, market cap, 52w high/low
• get_historical_data — historical OHLCV summary over a period (5d → max)
• technical_analysis — SMA20/50/200, RSI(14), volatility, momentum signals
• compare_stocks — rank multiple tickers by performance, volatility, RSI
• company_overview — sector, industry, P/E, dividend yield, beta, business summary
• generate_insight_report — combines everything into one plain-English report

Data comes from Yahoo Finance via the free yfinance library — no API key needed.

---

1. Install

cd financial-analyst-mcp
python3 -m venv venv
source venv/bin/activate        # Windows: venv\Scripts\activate
pip install -r requirements.txt

Quick sanity check (needs internet access to Yahoo Finance):

python server.py

It should just sit there waiting on stdio — that's correct, it's a stdio MCP server. Press Ctrl+C to stop. The real usage is through Claude Desktop or Cursor, which will launch it for you (next steps below).

---

2. Connect it to Claude Desktop

Open your Claude Desktop config file:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

Add (or merge into) the mcpServers block, using the absolute path to this project on your machine:

{
  "mcpServers": {
    "financial-analyst": {
      "command": "/absolute/path/to/financial-analyst-mcp/venv/bin/python",
      "args": ["/absolute/path/to/financial-analyst-mcp/server.py"]
    }
  }
}

On Windows, command would point to ...\\financial-analyst-mcp\\venv\\Scripts\\python.exe.

Restart Claude Desktop. You should see a small tools/plug icon indicating the financial-analyst server is connected, and Claude will be able to call its tools automatically when you ask stock-related questions.

---

3. Connect it to Cursor

In Cursor: Settings → MCP → Add new MCP server, and either paste the same JSON shape as above, or add it to .cursor/mcp.json in your project:

{
  "mcpServers": {
    "financial-analyst": {
      "command": "/absolute/path/to/financial-analyst-mcp/venv/bin/python",
      "args": ["/absolute/path/to/financial-analyst-mcp/server.py"]
    }
  }
}

Reload Cursor, then enable the server from the MCP panel.

---

4. Try it

Once connected, just ask naturally:

• "What's the current price of NVDA?"
• "Give me a technical analysis on TSLA over the last 3 months."
• "Compare AAPL, MSFT, and GOOGL performance this year."
• "Generate a full insight report on AMZN."

Claude will call the right tool(s) automatically and reason over the structured data it gets back.

---

Notes & next steps

• This is a read-only research tool — it doesn't place trades or move money anywhere.
• Reports are automated data summaries, not financial advice (the generate_insight_report tool says so explicitly).
• Yahoo Finance data can be delayed and occasionally rate-limits rapid requests — if a call fails, wait a few seconds and retry.
• Easy things to add later: caching, a get_news tool (via a news API), earnings calendar lookups, options chain data, or portfolio-level analysis across a basket of tickers.