MCP Server — Market Data Tools

MCP server providing real-time stock market data via Alpha Vantage API. Deploy to Railway for remote access from claude.ai scheduled tasks.

Tools

Tool	Description	Rate
get_full(ticker)	All data in one call (price, PE, EPS, RSI, SMA, MACD, analyst ratings)	6 API calls, ~75s
get_quote(ticker)	Current price, volume, change	1 call
get_overview(ticker)	Fundamentals (PE, EPS, margins, market cap, analyst ratings)	1 call
get_rsi(ticker, period)	RSI indicator (default 14-day)	1 call
get_sma(ticker, period)	Simple Moving Average (50 or 200)	1 call
get_macd(ticker)	MACD indicator	1 call
get_news(ticker)	News sentiment for a ticker	1 call
get_news_general()	General market news	1 call

Run locally

pip install -r requirements.txt
python server.py              # stdio (for Claude Code)
python server.py --remote     # HTTP (for remote access)

Deploy to Railway

1. Push this repo to GitHub
2. Go to railway.com → New Project → Deploy from GitHub
3. Set environment variable: ALPHA_VANTAGE_API_KEY (or use the default)
4. Railway gives you an HTTPS URL like https://mcp-server-xyz.railway.app

Connect to claude.ai

1. Go to claude.ai → Settings → Connectors → Add custom connector
2. Paste: https://your-railway-url.railway.app/mcp
3. Done — scheduled tasks can now call the tools

Connect to Claude Code (local)

claude mcp add market-data -- python /path/to/server.py

Rate limits

Alpha Vantage free tier: 5 API calls/minute, 25/day. get_full() uses 6 calls per ticker, so max ~4 tickers per day on free tier. Upgrade at alphavantage.co/premium for higher limits.