Alpaca Trading MCP

Alpaca Trading MCP

Skill by ara.so — MCP Skills collection.

Alpaca MCP Server is an official Model Context Protocol server that enables natural language trading operations through AI assistants. It provides comprehensive access to Alpaca's Trading API for stocks, options, crypto, portfolio management, and real-time market data. Built with FastMCP and OpenAPI, this is a complete v2 rewrite with spec-aligned tooling.

What It Does

• Market Data Access: Real-time and historical quotes, trades, bars for stocks, crypto, and options
• Trading Operations: Place market, limit, stop, trailing-stop orders for any supported asset
• Options Trading: Search contracts, place single/multi-leg strategies, get Greeks and IV
• Portfolio Management: View account info, positions, buying power, portfolio history
• Watchlist Management: Create, update, and manage custom watchlists
• News & Corporate Actions: Access market news and corporate action announcements

Installation

Prerequisites

• Python 3.10+
• uv package manager: curl -LsSf https://astral.sh/uv/install.sh | sh
• Alpaca API keys (free paper trading account at https://app.alpaca.markets)
• An MCP-compatible client (Claude Desktop, Cursor, VS Code, etc.)

Configuration

Add to your MCP client configuration file:

Claude Desktop (~/Library/Application Support/Claude/claude_desktop_config.json on Mac):

{
  "mcpServers": {
    "alpaca": {
      "command": "uvx",
      "args": ["alpaca-mcp-server"],
      "env": {
        "ALPACA_API_KEY": "your_alpaca_api_key",
        "ALPACA_SECRET_KEY": "your_alpaca_secret_key",
        "ALPACA_PAPER_TRADE": "true"
      }
    }
  }
}

Cursor (~/.cursor/mcp.json):

{
  "mcpServers": {
    "alpaca": {
      "command": "uvx",
      "args": ["alpaca-mcp-server"],
      "env": {
        "ALPACA_API_KEY": "your_alpaca_api_key",
        "ALPACA_SECRET_KEY": "your_alpaca_secret_key",
        "ALPACA_PAPER_TRADE": "true"
      }
    }
  }
}

VS Code (.vscode/mcp.json in project root):

{
  "mcp": {
    "servers": {
      "alpaca": {
        "type": "stdio",
        "command": "uvx",
        "args": ["alpaca-mcp-server"],
        "env": {
          "ALPACA_API_KEY": "your_alpaca_api_key",
          "ALPACA_SECRET_KEY": "your_alpaca_secret_key",
          "ALPACA_PAPER_TRADE": "true"
        }
      }
    }
  }
}

Claude Code CLI:

claude mcp add alpaca --scope user --transport stdio uvx alpaca-mcp-server \
  --env ALPACA_API_KEY=your_alpaca_api_key \
  --env ALPACA_SECRET_KEY=your_alpaca_secret_key \
  --env ALPACA_PAPER_TRADE=true

Environment Variables

Toolset Filtering

Restrict available tools by setting ALPACA_TOOLSETS:

{
  "env": {
    "ALPACA_TOOLSETS": "account,trading,stock-data"
  }
}

Available toolsets:

• account: Account info, portfolio history, activities
• trading: Orders, positions, options exercise
• watchlists: Watchlist CRUD operations
• assets: Asset lookup, option contracts, market calendar
• stock-data: Stock bars, quotes, trades, snapshots
• crypto-data: Crypto bars, quotes, trades, orderbooks
• options-data: Option bars, quotes, Greeks, chain
• corporate-actions: Corporate action announcements
• news: News articles

Core Trading Operations

Check Account Status

# Natural language prompts:
# "What's my current account balance?"
# "Show my buying power"
# "Am I approved for options trading?"

# The MCP server exposes get_account_info tool
# Returns: account_number, status, buying_power, cash, portfolio_value, etc.

Place Stock Orders

# Natural language prompts:
# "Buy 100 shares of AAPL at market price"
# "Place a limit order to buy TSLA at $250"
# "Sell 50 shares of NVDA with a trailing stop at 2%"

# Market order
# Tool: post_orders
# Parameters: {
#   "symbol": "AAPL",
#   "qty": 100,
#   "side": "buy",
#   "type": "market",
#   "time_in_force": "day"
# }

# Limit order
# Parameters: {
#   "symbol": "TSLA",
#   "qty": 10,
#   "side": "buy",
#   "type": "limit",
#   "limit_price": 250.00,
#   "time_in_force": "gtc"
# }

# Trailing stop
# Parameters: {
#   "symbol": "NVDA",
#   "qty": 50,
#   "side": "sell",
#   "type": "trailing_stop",
#   "trail_percent": 2.0,
#   "time_in_force": "gtc"
# }

Get Real-Time Market Data

# Natural language prompts:
# "What's the current price of AAPL?"
# "Show me the latest BTC quote"
# "Get today's price bars for MSFT in 5-minute intervals"

# Latest quote
# Tool: get_stocks_quotes_latest
# Parameters: {"symbols": "AAPL,MSFT,GOOGL"}

# Crypto quote
# Tool: get_crypto_quotes_latest
# Parameters: {"symbols": "BTC/USD,ETH/USD"}

# Historical bars
# Tool: get_stocks_bars
# Parameters: {
#   "symbols": "MSFT",
#   "timeframe": "5Min",
#   "start": "2025-05-01T09:30:00Z",
#   "end": "2025-05-01T16:00:00Z"
# }

Options Trading

# Natural language prompts:
# "Find call options for AAPL expiring next month"
# "Place a covered call on my 100 shares of NVDA"
# "Get the Greeks for SPY 500 calls expiring this Friday"

# Search option contracts
# Tool: get_options_contracts
# Parameters: {
#   "underlying_symbols": "AAPL",
#   "expiration_date_gte": "2025-06-01",
#   "expiration_date_lte": "2025-06-30",
#   "type": "call"
# }

# Place option order
# Tool: post_orders
# Parameters: {
#   "symbol": "AAPL250620C00150000",
#   "qty": 1,
#   "side": "sell_to_open",
#   "type": "limit",
#   "limit_price": 5.50,
#   "time_in_force": "day",
#   "class": "option"
# }

# Get option snapshot with Greeks
# Tool: get_options_snapshots
# Parameters: {"symbols": "SPY250516C00500000"}
# Returns: greeks (delta, gamma, theta, vega), implied_volatility, latest_quote

Manage Positions

# Natural language prompts:
# "Show all my open positions"
# "What's my position in AAPL?"
# "Close half of my TSLA position"

# Get all positions
# Tool: get_positions

# Get specific position
# Tool: get_positions_by_symbol_or_asset_id
# Parameters: {"symbol_or_asset_id": "AAPL"}

# Close position
# Tool: delete_positions_by_symbol_or_asset_id
# Parameters: {
#   "symbol_or_asset_id": "TSLA",
#   "percentage": 50  # Optional: close percentage
# }

Watchlist Management

# Natural language prompts:
# "Create a tech stocks watchlist with AAPL, MSFT, GOOGL"
# "Add TSLA to my tech watchlist"
# "Show me all my watchlists"

# Create watchlist
# Tool: post_watchlists
# Parameters: {
#   "name": "Tech Stocks",
#   "symbols": ["AAPL", "MSFT", "GOOGL"]
# }

# Get all watchlists
# Tool: get_watchlists

# Add symbol to watchlist
# Tool: post_watchlists_by_watchlist_id
# Parameters: {
#   "watchlist_id": "uuid-here",
#   "symbol": "TSLA"
# }

Advanced Market Data Queries

Screen Stocks

# Natural language prompts:
# "Find the most active stocks today"
# "Show me top gainers with volume over 1M"

# Tool: get_screener_stocks_most_actives
# Parameters: {
#   "by": "volume",
#   "top": 10
# }

# Tool: get_screener_stocks_movers
# Parameters: {
#   "market_type": "stocks",
#   "top": 20
# }

Get News

# Natural language prompts:
# "Show me latest news about AAPL"
# "Get crypto news for Bitcoin"

# Stock news
# Tool: get_news
# Parameters: {
#   "symbols": "AAPL",
#   "limit": 10,
#   "sort": "desc"
# }

# Crypto news
# Tool: get_crypto_news
# Parameters: {
#   "symbols": "BTC",
#   "limit": 10
# }

Corporate Actions

# Natural language prompts:
# "Are there any upcoming stock splits?"
# "Show dividend announcements for next week"

# Tool: get_corporate_actions_announcements
# Parameters: {
#   "ca_types": "dividend,split",
#   "since": "2025-05-01",
#   "until": "2025-05-31"
# }

Portfolio Analysis

Get Portfolio History

# Natural language prompts:
# "Show my portfolio performance over the last month"
# "Graph my portfolio value for the past year"

# Tool: get_account_portfolio_history
# Parameters: {
#   "period": "1M",
#   "timeframe": "1D"
# }
# Returns: timestamp, equity, profit_loss, profit_loss_pct arrays

Account Activities

# Natural language prompts:
# "Show my recent trades"
# "List all dividend payments this month"

# Tool: get_account_activities
# Parameters: {
#   "activity_types": "FILL,DIV",
#   "after": "2025-05-01T00:00:00Z",
#   "direction": "desc"
# }

Common Patterns

Multi-Symbol Quotes

# Get quotes for multiple symbols at once
# "Get current prices for AAPL, MSFT, GOOGL, AMZN"
# Tool: get_stocks_quotes_latest
# Parameters: {"symbols": "AAPL,MSFT,GOOGL,AMZN"}
# Use feed parameter for realtime: "feed": "sip"

Batch Order Placement

# Place multiple orders together
# "Buy 10 shares each of AAPL, MSFT, GOOGL at market"
# Call post_orders tool 3 times or use a loop pattern
# Each order is submitted individually through the API

Conditional Order Logic

# Complex strategies require multiple steps
# "If AAPL drops below $150, buy 100 shares"
# 1. Monitor current price with get_stocks_quotes_latest
# 2. When condition met, call post_orders
# Note: Alpaca supports bracket/OCO orders natively:
# Use order_class: "bracket" with take_profit and stop_loss

Options Strategies

# Multi-leg options spreads
# "Create a bull call spread on SPY"
# Tool: post_orders with order_class: "multileg"
# Parameters: {
#   "symbol": "SPY",
#   "type": "limit",
#   "time_in_force": "day",
#   "order_class": "multileg",
#   "legs": [
#     {"symbol": "SPY250620C00500000", "side": "buy_to_open", "qty": 1},
#     {"symbol": "SPY250620C00510000", "side": "sell_to_open", "qty": 1}
#   ]
# }

Troubleshooting

Server Not Showing in MCP Client

1. Restart your MCP client after editing config
2. Check JSON syntax in your config file (trailing commas cause issues)
3. Verify uv is installed: uv --version
4. Check logs in client's developer console/logs

Authentication Errors

Error: 401 Unauthorized

• Verify ALPACA_API_KEY and ALPACA_SECRET_KEY are correct
• Ensure you're using paper trading keys if ALPACA_PAPER_TRADE=true
• Check if keys have necessary permissions in Alpaca dashboard

Tool Not Found

Error: Tool 'old_tool_name' not found

• You may be using v1 tool names. V2 is a complete rewrite.
• Clear MCP client cache and restart
• Start a fresh chat session
• Check Available Tools for current tool names

Rate Limiting

Error: 429 Too Many Requests

• Alpaca API has rate limits (200 requests/minute for paper trading)
• Batch symbol queries when possible (comma-separated symbols)
• Add delays between rapid tool calls
• Upgrade to live account for higher limits

Order Rejected

Error: insufficient buying power / position limit exceeded

• Check account buying power: get_account_info
• Verify position limits haven't been hit
• For options, ensure account has appropriate trading level
• Check if market is open: get_clock

Market Data Gaps

Error: no data found for timeframe

• Market may be closed (check get_clock)
• Symbol may not have data for requested timeframe
• Use feed: "sip" for consolidated US market data
• Crypto trades 24/7, stocks only during market hours (9:30-16:00 ET)

V1 to V2 Migration

If upgrading from v1:

1. Remove any .env files — v2 uses env vars in MCP config only
2. Delete v1 tool references in custom instructions/rules
3. Restart MCP client to clear tool cache
4. Start fresh session so LLM discovers new tools
5. Update toolset filtering using ALPACA_TOOLSETS env var

Docker Issues

# If server fails to start in Docker
docker logs <container-id>

# Check environment variables are passed
docker run --rm -e ALPACA_API_KEY=test mcp/alpaca:latest env | grep ALPACA

# Test connection manually
docker run --rm -i \
  -e ALPACA_API_KEY=your_key \
  -e ALPACA_SECRET_KEY=your_secret \
  -e ALPACA_PAPER_TRADE=true \
  mcp/alpaca:latest

Live Trading

To switch from paper to live trading:

{
  "env": {
    "ALPACA_API_KEY": "your_live_api_key",
    "ALPACA_SECRET_KEY": "your_live_secret_key",
    "ALPACA_PAPER_TRADE": "false"
  }
}

⚠️ WARNING: Live trading uses real money. Test thoroughly in paper trading first. Markets are volatile and losses can exceed expectations.

Additional Resources

• Alpaca API Docs: https://docs.alpaca.markets/
• Paper Trading Dashboard: https://app.alpaca.markets/paper/dashboard/overview
• Live Trading Dashboard: https://app.alpaca.markets/brokerage/dashboard/overview
• Python SDK: https://alpaca.markets/sdks/python/
• MCP Specification: https://modelcontextprotocol.io/
• Community Forum: https://forum.alpaca.markets/
• Slack Community: https://alpaca.markets/slack

License

MIT License - Alpaca Markets