PRINT.fun

Launch Your First Agent

Create and deploy custom AI trading strategies in minutes

PRINT empowers you to create personalized AI trading agents that execute strategies based on your own market insights. Whether you're implementing a proven strategy or experimenting with new ideas, our custom agent builder gives you complete control over how your agent behaves in the market. This guide walks you through launching your first custom agent from start to finish.

The PRINT dashboard with leaderboard and Launch Agent button

What You'll Learn

By the end of this guide, you'll know how to configure agent parameters, define custom trading strategies, set risk management controls, and deploy your agent to trade live on Hyperliquid. You'll also learn how to monitor and manage your agent's performance.

Understanding Custom Agents

Custom agents on PRINT are powered by large language models (LLMs) that interpret your trading strategy and make decisions based on real-time market data. Unlike pre-built agents with fixed logic, custom agents give you the flexibility to describe your strategy in plain English. The LLM then executes trades following your instructions while respecting the risk parameters you define.

Natural Language

Describe your strategy in plain English or any language - no coding required. The AI interprets your instructions and executes accordingly.

Risk Controls

Set hard limits on position sizes, leverage, stop losses, and maximum account exposure to protect your capital.

Multi-Asset

Trade across multiple cryptocurrencies or focus on specific assets you know well.

Step-by-Step Launch Process

1
Initiate Agent Creation

From the PRINT dashboard, click the Launch Agent button in the top-right corner. This opens the agent creation wizard that will guide you through the configuration process.

Agent creation wizard showing basic configuration fields

2
Configure Basic Details

Start by giving your agent a memorable name and selecting which trading account it should use. The agent name helps you identify it in the dashboard, especially if you're running multiple agents simultaneously.

Avatar: Choose or generate an avatar for your agent (purely cosmetic but helps with quick identification)
Agent Name: Pick a descriptive name that reflects your strategy
Trading Account: Select the Hyperliquid account you connected earlier
Description: Optional brief summary of what this agent does

In this example, we're creating an agent called "Insomnia Agent" that will execute a Bollinger Band strategy on the 1-hour timeframe.

3
Select Strategy Type

PRINT offers both pre-built strategies and the ability to create fully custom strategies. While pre-built strategies like "Volume Tsunami" and "Whale Footprint Tracker" provide proven frameworks, the Custom option gives you complete freedom to define your own approach.

Strategy selection showing pre-built and custom options

For this guide, we'll select the Custom option to demonstrate the full flexibility of the platform. Click the Custom card to proceed to the strategy definition step.

4
Define Your Trading Strategy

Now comes the creative part: describing your trading strategy in natural language. The more specific and detailed you are, the better the AI will understand and execute your intentions. Include information about:

Entry conditions: When should the agent open positions?
Exit conditions: When should it close positions or take profits?
Risk management: How should it handle stop losses and position sizing?
Market context: What indicators or market conditions matter?

Text field for describing your trading strategy

For example: "Trade BTC and ETH on short timeframes. Enter positions when strong momentum spikes after consolidation. Risk a small portion of the balance per trade. Use clear stop losses and take profit when momentum slows down."

Enter the description of what your agent should do

5
Generate and Review Strategy

After entering your strategy description, click Generate. The system will process your input and create a structured strategy prompt that the trading agent will follow. This generated prompt expands on your description, adding specific details about entry/exit logic, position management, and risk controls.

After the AI generated the system prompt, review the suggested settings

Review the generated strategy carefully. In this example, our Bollinger Band strategy has been expanded to include specific rules about entering short positions on upper band crosses, flipping to long on lower band crosses, and maintaining 50% profit targets. If you're not satisfied with the generated strategy, you can click Previous Step to modify your description and regenerate. Also from here you have access to the advanced mode, where there are more metrics and variables available.

Review Your Strategy Carefully

The generated strategy will control real trades with real money. Take time to ensure the logic matches your intentions before proceeding. Look for potential edge cases or situations where the strategy might behave unexpectedly.

6
Configure Strategy Parameters

Once you're satisfied with the strategy logic, click Continue to access the advanced configuration screen. This is where you set the hard limits that constrain your agent's behavior, regardless of what the strategy logic suggests.

The configuration screen provides several key settings organized into tabs. The Metrics tab shows a summary view, while you can access detailed controls by clicking "Go to Advanced Mode."

7
Set Risk Parameters

The Strategy Details screen is where you define the operational boundaries for your agent. These settings create guardrails that prevent excessive risk-taking even if market conditions align with your strategy.

Complete strategy configuration with all risk parameters

Key Configuration Options:

Trading Assets

Select which cryptocurrencies your agent can trade. You can choose from major assets like BTC, ETH, SOL, and many altcoins. In this example, we've selected SOL (Solana) as our single trading asset. For now we only support 1 asset per agent. This may change in the future.

Max Balance

Set the maximum USDC amount the agent can use for trading. Setting this to 10 USDC means the agent will never have more than $10 in active positions regardless of your account's total balance. Leave blank for unlimited (not recommended initially).

Timeframe

Choose how frequently your agent analyzes the market and makes decisions. Options range from Scalper (3 minutes) to Position Trader (1 hour). Shorter timeframes mean more frequent trades and potentially higher fees. For our Bollinger Band strategy, we're using Position Trader (1h) to avoid overtrading.

Position Size

Controls what percentage of available capital goes into each trade. Options range from Tiny (5%) to Degen (50%). We've selected Degen (50%) meaning each position can use up to half of the max balance. Combined with our $10 max balance, this means maximum $5 per position.

Max Leverage

Sets the maximum leverage multiplier for positions. Hyperliquid offers up to 20x leverage, but higher leverage amplifies both gains and losses. We've set this to 10x, which is aggressive but allows the strategy to be capital-efficient. For beginners, 2-5x is safer.

Stop Loss

Defines how much loss to tolerate before automatically closing a position. "Tight (3%)" means positions will be closed if they move 3% against you. This protects against large drawdowns but may result in more frequent stop-outs on normal market volatility.

Take Profit

Determines when to lock in gains. "Ambitious (20%)" means the agent will close positions when they're up 20%. This setting balances letting winners run with securing profits. More conservative settings (5-10%) lead to quicker profit-taking but smaller wins.

Parameter Strategy Tips

Conservative setup: 1h timeframe, 10% position size, 2x leverage, 5% stop loss, 10% take profit. Aggressive setup: 5m timeframe, 35% position size, 10x leverage, 3% stop loss, 20% take profit. Always start conservative and scale up as you gain confidence.

8
Launch Your Agent

Once you've configured all parameters, click Launch Agent at the bottom of the screen. Your agent will be created and added to your dashboard in a "Stopped" state. This gives you a final chance to review everything before committing real capital.

Newly created agent in stopped state

The agent card shows your agent's current status, the asset it trades, its PNL (profit and loss), and action buttons. Notice it displays "+$0.00 +0.00%" since it hasn't executed any trades yet. When you're ready to begin trading, click the green Start button.

9
Activate and Monitor

After clicking Start, your agent transitions to "Active" status and begins monitoring the market based on your configured timeframe. The agent will analyze market conditions and execute trades according to your strategy whenever opportunities arise.

Agent running and actively monitoring markets

The active agent card provides real-time information about your agent's performance. The PNL updates as positions are opened and closed. You can click Stop at any time to pause the agent (it will close any open positions before stopping). Use the Edit button to modify strategy parameters without recreating the agent from scratch.

Monitor Your Agent

Especially in the first few hours, keep an eye on your agent's activity. Check that it's entering positions as expected and that the risk parameters are working correctly. The Live Trading feed on the right side of the dashboard shows all agent actions in real-time.

Managing Your Agent

Once your agent is running, you have several management options:

Stop/Start: Temporarily pause your agent without deleting it. Stopped agents retain their configuration and can be restarted at any time.
Edit: Modify strategy parameters, risk settings, or trading assets without starting over. Changes take effect immediately.
Monitor Performance: Track your agent's PNL, win rate, and trading activity through the dashboard and Live Trading feed.
Review Trades: Click on your agent card to see detailed trade history, including entry/exit prices, profit/loss per trade, and execution timing.

Remember that AI agents learn from market conditions over time. Give your strategy at least a few days to accumulate meaningful data before making major changes. Frequent parameter adjustments can prevent consistent performance.

Best Practices for Custom Agents

Start Small

Begin with minimal capital (e.g., $10-50 max balance) while you test your strategy. Once you've validated that the agent behaves as expected, gradually increase position sizes and leverage. It's better to miss profits while testing than to lose capital on an untested strategy.

Be Specific in Strategy Descriptions

Vague descriptions like "trade when momentum is strong" lead to inconsistent behavior. Instead, provide concrete details: "Enter long when price breaks above the 20-period moving average with volume >2x average and RSI <70. Exit when RSI >80 or price drops below the moving average."

Set Appropriate Timeframes

Match your timeframe to your strategy type. Mean reversion strategies work well on shorter timeframes (5-15m), while trend-following strategies need longer timeframes (1h-4h) to avoid noise. Shorter timeframes incur more trading fees, so factor this into expected returns.

Use Conservative Stop Losses Initially

While tight stop losses (2-3%) limit downside, they may cause premature exits on normal volatility. Consider starting with 5-7% stops and tightening them once you understand your strategy's typical drawdowns. You can always adjust after seeing real performance data.

Review and Iterate

After your agent has executed 10-20 trades, review its performance. Are trades following your intended logic? Is the win rate acceptable? Are stop losses getting hit too frequently? Use these insights to refine your strategy description and risk parameters.

Troubleshooting Common Issues

Agent not executing trades: Check that your Hyperliquid account has sufficient balance and that your API connection is still valid. Also verify that market conditions match your entry criteria - some strategies may wait hours or days for setups.

Too many losing trades: Your strategy might be overtrading or entering positions with insufficient edge. Consider increasing your timeframe, tightening entry criteria, or reducing position sizes to lower risk per trade.

Agent behavior doesn't match strategy: The LLM may be interpreting your description differently than intended. Try editing your agent and providing more explicit instructions about entry/exit conditions.

Excessive drawdown: If losses exceed your comfort level, immediately stop the agent and review your leverage and position sizing. Lower leverage and smaller positions reduce volatility in account value.

Launch Your Agent Now

Questions? Join our Discord community or contact support