Market Data Skill

This skill provides patterns for fetching and managing market data in the wicktrade system.

Two Ways to Access Data

1. Alpaca MCP Server (Recommended for Interactive Use)

The Alpaca MCP server is configured and provides direct access to market data tools:

Available MCP Tools:

get_stock_bars - Historical OHLCV data
get_stock_quotes - Real-time quotes
get_stock_snapshot - Current market snapshot
get_stock_trades - Trade-level data
get_crypto_bars - Crypto historical data
get_crypto_quotes - Crypto quotes
is_market_open - Check market status
get_calendar - Trading calendar

Example MCP Usage:

Use get_stock_bars to fetch SPY data for the last 30 days at 15Min timeframe

2. Python DataManager (For Backtesting)

from wicktrade.core.data_manager import DataManager
from datetime import datetime, timedelta

dm = DataManager()
candles = dm.get_candles(
    symbol="SPY",
    timeframe="15Min",
    start=datetime.now() - timedelta(days=30),
)

Supported Timeframes

Timeframe	MCP Arg	Python Arg
1 minute	1Min	"1Min"
5 minutes	5Min	"5Min"
15 minutes	15Min	"15Min"
30 minutes	30Min	"30Min"
1 hour	1Hour	"1Hour"
4 hours	4Hour	"4Hour"
1 day	1Day	"1Day"

MCP Server Patterns

Get Historical Bars

Use get_stock_bars for AAPL with:
- timeframe: 15Min
- start: 2024-01-01
- limit: 1000

Get Real-Time Quote

Use get_stock_quotes for SPY to get current bid/ask

Check Market Status

Use is_market_open to check if trading is available

Get Trading Calendar

Use get_calendar to see upcoming market hours

Python Patterns

Download Data for Multiple Symbols

dm = DataManager()
symbols = ["SPY", "QQQ", "AAPL", "MSFT"]
results = dm.download_symbols(symbols, timeframe="15Min")

Check Cache Status

cache_info = dm.get_cache_info()
for info in cache_info:
    print(f"{info['symbol']} {info['timeframe']}: {info['rows']} bars")

Clear Cache

dm.clear_cache(symbol="SPY")  # Specific symbol
dm.clear_cache()               # All cached data

Data Validation

# Check for gaps in data
candles = dm.get_candles("SPY", "15Min")
for i in range(1, len(candles)):
    gap = (candles[i].timestamp - candles[i-1].timestamp).total_seconds()
    expected = 15 * 60
    if gap > expected * 2:
        print(f"Gap detected at {candles[i].timestamp}")

When to Use Each Method

Use Case	Recommended Method
Quick data check	MCP Server
Backtesting	Python DataManager
Real-time quotes	MCP Server
Bulk historical download	Python DataManager
Market status check	MCP Server
Cached data access	Python DataManager

market-data