DEV Community

vdalhambra
vdalhambra

Posted on

How to add real-time stock data to Claude in under 2 minutes

#claudeai #ai #python #mcp

If you've tried asking Claude about a stock price, you've probably seen this frustrating response:

"I don't have real-time data. As of my training cutoff, AAPL was around..."

Your training data is always stale. Markets move every second. You need real-time data.

Here's how to fix that in under 2 minutes using FinanceKit MCP — an MCP (Model Context Protocol) server that gives any AI agent 17 financial analysis tools.

What you can ask Claude after installing this

  • "What's AAPL's current price and RSI?"
  • "Run full technical analysis on NVDA" → Gets RSI, MACD, Bollinger Bands, Golden Cross detection, plain-English verdict
  • "Compare TSLA vs GM vs F over the last 6 months" → Returns, volatility, Sharpe ratios, max drawdown
  • "What's the best-performing sector this month?" → Ranks all 11 GICS sectors
  • "Analyze my portfolio: AAPL:10, MSFT:5, NVDA:3" → Full breakdown with concentration risk
  • "What's Bitcoin doing? Show RSI and trend" → Technical analysis works for crypto too

The 2-minute install

Option 1: MCPize (zero setup — recommended)

If you don't want to mess with Python or terminal commands, this is the easiest:

  1. Go to mcpize.com/mcp/financekit-mcp
  2. Click "Install in Claude Desktop" (or Cursor, Windsurf, Claude Code)
  3. Done. Free tier includes 100 calls/month.

Or add this to your MCP config directly:

{
  "mcpServers": {
    "financekit": {
      "url": "https://financekit-mcp.mcpize.run/mcp"
    }
  }
}
Enter fullscreen mode Exit fullscreen mode

Option 2: Self-hosted via uvx (for developers) 

# Claude Code
claude mcp add financekit -- uvx --from financekit-mcp financekit
Enter fullscreen mode Exit fullscreen mode

Or in claude_desktop_config.json:

{
  "mcpServers": {
    "financekit": {
      "command": "uvx",
      "args": ["--from", "financekit-mcp", "financekit"]
    }
  }
}
Enter fullscreen mode Exit fullscreen mode

Requires uv (Python package manager): pip install uv or brew install uv.

What makes this different from a simple Yahoo Finance wrapper

Most MCP servers just forward API responses. FinanceKit processes the data:

Technical analysis with a verdict

Ask "Run technical analysis on NVDA" and instead of raw numbers, you get:

Symbol: NVDA
Current Price: 142.58
Overall bias: STRONGLY BULLISH (3.0 bullish vs 0.5 bearish signals)

Indicators:
  RSI(14): 55.65 — neutral
  MACD: histogram positive — bullish momentum
  Bollinger Bands: price within bands — normal
  SMA(50): 138.84 → price above SMA50 = uptrend
  ADX: 28.3 — strong trend

Patterns:
  Golden Cross: YES (SMA50 crossed SMA200 upward)
  Overbought: false
  Oversold: false

Signals:
  → "MACD bullish — line above signal, both positive. Strong upward momentum."
  → "ADX > 25 indicates strong directional trend."
  → "Price above all SMAs (20/50/200) confirms uptrend."
Enter fullscreen mode Exit fullscreen mode

That's 10 calculated indicators + pattern detection + plain-English signals. You can't get that from a raw API call.

Risk metrics that matter

Ask "What's the risk profile of TSLA over the last year?":

Symbol: TSLA
Metrics:
  annual_return_pct: 22.45
  annual_volatility_pct: 48.2
  sharpe_ratio: 0.87
  sortino_ratio: 1.21
  beta: 1.85 (vs SPY)
  value_at_risk_95_pct: -4.8
  max_drawdown_pct: -32.4

Risk rating: MODERATE
Enter fullscreen mode Exit fullscreen mode

Sharpe, Sortino, Beta, VaR, Max Drawdown — all computed from 252 trading days.

Portfolio analysis with sector breakdown

"Analyze my portfolio: AAPL:10, MSFT:5, NVDA:3, GOOGL:2, AMZN:1":

Total Value: $8,412.30
Positions: 5
Concentration Risk: MODERATE (top holding 31% of portfolio)

Sector Breakdown:
  Technology: 68.2%
  Communication Services: 22.4%
  Consumer Discretionary: 9.4%
Enter fullscreen mode Exit fullscreen mode

All 17 tools

Stocks (3): stock_quote, multi_quote, company_info
Crypto (4): crypto_price, crypto_trending, crypto_search, crypto_top_coins
Technical (2): technical_analysis, price_history
Market (1): market_overview (S&P 500, NASDAQ, DOW, VIX, movers)
Comparison (2): compare_assets, portfolio_analysis
Premium (5): risk_metrics, correlation_matrix, earnings_calendar, options_chain, sector_rotation

Pricing (MCPize hosted)

Tier Price Calls/month For who
Free $0 100 Testing
Hobby $9/mo 2,500 Personal investing
Pro $29/mo 10,000 Active traders
Team $79/mo 50,000 Small funds
Business $179/mo 200,000 Serious ops

Self-hosted is always free (just pip install), but you manage infra and Yahoo Finance rate limits yourself.

Pair with SiteAudit MCP for full-stack analysis

For web/business audits on top of financial analysis, pair with SiteAudit MCP (11 tools for SEO/security/performance/accessibility). Bundle is $39/mo for both Pro tiers.

Data sources

  • Stocks: Yahoo Finance (no API key needed)
  • Crypto: CoinGecko free tier (10K calls/month)
  • Technicals: Calculated locally using ta library

All cached (quotes 60s, historical 1h, crypto 2min) to minimize calls.

Links

If this saves you time, star the repo on GitHub — it helps more devs find it.

Top comments (0)