Data source: Historical pattern similarity from Chart Library for signal generation

[grahammccain]

Summary

OpenAlice is a file-driven AI trading agent — it reads market data and generates trading signals. Historical chart pattern similarity could be a useful input signal: "when the chart looked like this before, what happened next?"

Chart Library provides this via API and MCP server. It has 24M+ pre-computed pattern embeddings across 19K US equities (10 years of minute-bar data).

How it could integrate with OpenAlice

Since OpenAlice is file-driven, the simplest integration would be a data fetcher that writes pattern similarity data to a file:

import requests
import json

def fetch_pattern_data(symbols: list[str], date: str, output_file: str):
    """Fetch pattern similarity data and write to file for OpenAlice."""
    results = {}
    for symbol in symbols:
        resp = requests.get("https://chartlibrary.io/api/v1/search", params={
            "symbol": symbol, "date": date, "timeframe": "RTH"
        }, headers={"X-API-Key": "your-key"})
        
        matches = resp.json()["matches"]
        results[symbol] = {
            "date": date,
            "top_match_distance": matches[0]["distance"],
            "avg_5d_return": sum(m["return_5d"] for m in matches) / len(matches),
            "win_rate_5d": sum(1 for m in matches if m["return_5d"] > 0) / len(matches),
            "avg_10d_return": sum(m["return_10d"] for m in matches) / len(matches),
            "matches": [{
                "symbol": m["symbol"], "date": m["date"],
                "distance": m["distance"], "return_5d": m["return_5d"]
            } for m in matches[:5]]
        }
    
    with open(output_file, "w") as f:
        json.dump(results, f, indent=2)

For AI agent mode (MCP):

pip install chartlibrary-mcp

19 MCP tools including search_pattern, get_risk_adjusted_picks, get_regime_win_rates, and get_exit_signal.

What makes this useful for trading agents

• Base rate prior: Before making a prediction, see what historically happened after similar patterns
• Regime awareness: API filters by market regime (VIX level, etc.)
• Anomaly detection: /api/v1/anomaly/{symbol} flags unusual price action

Details

• Free tier: 200 calls/day
• Docs: https://chartlibrary.io/developers
• PyPI: pip install chartlibrary-mcp
• Coverage: 19K US equities, 8 timeframes, 10 years