Quick Start

This guide gets you from zero to a backtested AI trading agent in under 15 minutes. You’ll set up your environment, write a simple trading agent using Backtrader, upload it to our platform for backtesting, and view your performance metrics and tier badge.

Prerequisites

You’ll need basic Python knowledge and a Spooky Labs account. Don’t have an account yet? Join the waitlist.

Step 1: Create Your Agent

Create a file named agent.py with the following code:

import backtrader as bt
from chronos import ChronosPipeline
import torch

class Agent(bt.Strategy):
    params = (
        ('lookback', 20),
        ('model_name', 'amazon/chronos-t5-small'),
    )

    def __init__(self):
        self.forecaster = ChronosPipeline.from_pretrained(
            self.params.model_name,
            device_map="auto",
            torch_dtype=torch.float32
        )

    def prenext(self):
        self.next()

    def next(self):
        for d in self.datas:
            if len(d) < self.params.lookback:
                continue

            prices = list(d.close.get(size=self.params.lookback))
            context = torch.tensor(prices).unsqueeze(0)

            forecast = self.forecaster.predict(context, prediction_length=1)
            predicted = forecast[0].mean().item()

            pos = self.getposition(d).size
            if not pos and predicted > d.close[0]:
                size = int(self.broker.getcash() * 0.15 / d.close[0])
                self.buy(data=d, size=size)
            elif pos and predicted < d.close[0]:
                self.close(data=d)

Important requirements:

Your agent must follow these requirements exactly or the backtest will fail:

File must be named exactly agent.py
Class must be named exactly Agent
Must extend bt.Strategy
Must implement __init__(), prenext(), and next()
prenext() must call self.next() to avoid waiting for all indicator data

Step 2: Upload via Console

Now that you have your agent file, upload it to the platform. We’ll automatically backtest it against 2 years of historical market and news data from 2024 and 2025, with other data types coming soon, and generate performance metrics.

Go to your Console
Click the Upload Agent section
Select your agent.py file
Click Upload Agent

The platform will store your file securely, start a backtest automatically, show results in the Console once complete, and automatically rank your agent on the Leaderboard.

Step 3: View Your Results

Once the backtest completes, your agent card in the Console will show key performance metrics:

Metric	Description
Portfolio Return	Total percentage return on $10,000 starting capital
Sharpe Ratio	Risk-adjusted return (higher = better)
Win Rate	Percentage of profitable trades
Tier Badge	Unranked / Bronze / Silver / Gold based on return %

Click View Details to see the full performance breakdown including max drawdown, trade count, and profit factor.

What’s Next?

Your agent passed backtesting? Great! Now you can:

Deploy to Paper Trading — Trade with live market data and $10,000 virtual capital
Learn More About Writing Agents — Explore indicators, order types, and advanced patterns
Understand Your Metrics — Learn what makes a good agent

Quick Reference

Agent Requirements

Requirement	Value
Filename	`agent.py` (exactly)
Class name	`Agent` (exactly)
Base class	`bt.Strategy`
Required methods	`__init__()`, `prenext()`, `next()`

Status Flow

The status of your agent, shown in both the Console and Agent Detail pages, follows this diagram:

flowchart LR A[stored] --> B[building] B --> C[success] B --> D[failed] style C fill:#10b981,color:#fff; style D fill:#ef4444,color:#fff;

Troubleshooting

Backtest failed? Check the following:

Issue	Solution
Wrong filename	Your file must be named exactly `agent.py`.
Wrong class name	Your class must be named exactly `Agent`.
Missing prenext	You must implement `def prenext(self): self.next()`.
Syntax errors	Ensure there are no syntax errors in your Python code.

See Troubleshooting for more help.