This is the tutorial collection you bookmark and come back to. Whether you're setting up your first AI trading assistant or building a fully automated multi-strategy pipeline, each tutorial below builds on the previous one. Follow them in order, or jump to the section that matches your current skill level.
Difficulty: Beginner | Time: 15 minutes
Every trading system needs a reliable foundation. For OpenClaw, that means a cloud server that runs 24/7 without babysitting.
Steps:
What you'll have: A running OpenClaw agent on reliable cloud infrastructure, ready for trading skills.
Difficulty: Beginner | Time: 10-20 minutes
Your agent needs a way to talk to you. Telegram is the recommended starting point for traders — it's fast, supports rich formatting, and works great on mobile.
Steps:
What you'll have: A direct communication line between you and your trading agent, accessible from any device.
Difficulty: Beginner | Time: 15 minutes
An agent without data is just a chatbot. The skill installation guide walks through the process of adding capabilities to your OpenClaw instance.
Steps:
What you'll have: An agent that can pull real-time and historical market data on demand.
Difficulty: Beginner | Time: 20 minutes
Now let's make the agent actually useful. Price alerts are the simplest form of trading automation and the foundation everything else builds on.
Configuration approach:
Tell your agent what to watch:
"Monitor these stocks: AAPL, MSFT, NVDA, GOOGL, AMZN. Alert me immediately if any of them moves more than 3% from today's open."
The agent will:
Enhancements to try:
Difficulty: Intermediate | Time: 30 minutes
Move beyond price alerts to systematic technical screening.
Example: Oversold Bounce Screen
Configure your agent to run this screen daily at market close:
"Screen all S&P 500 stocks. Find stocks where RSI(14) is below 30, the price is above the 200-day SMA, and today's volume is at least 1.5x the 20-day average. Send results to my Telegram."
Example: Momentum Breakout Screen
"Find stocks making new 52-week highs with volume at least 2x the 20-day average. Exclude stocks with market cap below $1 billion."
Key learning: At this stage, you're defining systematic rules that the agent applies consistently. No more "I felt like the chart looked good" — every signal has a quantifiable basis.
Difficulty: Intermediate | Time: 45 minutes
Combine multiple indicators into a composite signal.
Example: Trend-Following Entry System
Define a scoring system:
Signal rule: Alert when a stock scores 4 or 5 out of 5.
Configure your agent to calculate this composite score for your watchlist daily and rank stocks by score. Push the ranked list to Discord for team review or Telegram for personal use.
Difficulty: Advanced | Time: 30 minutes
Good entries with bad risk management still lose money. Automate the risk calculations.
Position Sizing:
"For each trade signal, calculate the recommended position size based on: account size $50,000, maximum risk per trade 1.5%, and stop-loss distance based on 2x ATR(14)."
Portfolio Exposure Monitoring:
"Track my current positions. Alert me if any single position exceeds 10% of portfolio value, or if any sector represents more than 30% of total exposure."
Drawdown Alerts:
"Monitor my portfolio value daily. Alert me if the drawdown from the peak exceeds 5%, 10%, or 15%."
Difficulty: Advanced | Time: 1-2 hours
This is where everything comes together. Build a complete pipeline:
Each step is a scheduled task within OpenClaw, chaining skills together in a defined sequence.
Difficulty: Advanced | Time: 30 minutes
Set up automated performance tracking:
Use this data to refine your strategy parameters. If your RSI threshold of 30 generates too many false signals, try 25. If your holding period of 10 days is too short, test 15. Let the data guide your optimization.
As your trading system grows in complexity, you may need to upgrade your server. The Tencent Cloud Lighthouse Special Offer includes higher-tier instances for demanding workloads. The upgrade path is smooth — snapshot your current instance, spin up a larger one, restore, and you're running.
| Tutorial | Level | Key Skill Unlocked |
|---|---|---|
| 1-3 | Beginner | Infrastructure + Data |
| 4-5 | Beginner-Intermediate | Alerts + Screening |
| 6 | Intermediate | Multi-Indicator Strategies |
| 7 | Advanced | Risk Management |
| 8-9 | Advanced | Full Automation + Analytics |
Don't skip levels. Each tutorial builds muscle memory and confidence that the next one depends on. The traders who succeed with AI-assisted trading are the ones who validated at every step before adding complexity. Start with Tutorial 1 today.