Trading agents

AI Crypto Trading Agents with Defined Permissions

Create an agent in TRUE, choose its strategy and allocation, and monitor activity within the permissions you grant.

Create an agent →How to get started ↓

01 / Trading agents

How it works

Agents monitor conditions and execute within their configuration. Use the workflows available in Agent Studio, or create a supported draft through MCP. Autopilot provides a dedicated guided setup for automated market monitoring and trading.

Trading agents

Plan

ScopeAllocation

Review

PermissionsConfirmation

Follow through

ExecutionActivity

Illustrative walkthrough · no live data or orders

02 / Trading agents

How to get started

  1. 01

    Choose the workflow

    Open Agent Studio and select the type of agent available for your account.

  2. 02

    Define its scope

    Set the budget and review the strategy, assets and available controls.

  3. 03

    Fund and authorize

    Follow the app’s funding and wallet-approval steps. A saved draft is not an active agent.

  4. 04

    Inspect its activity

    Check the live status, positions and execution history in My Agents.

What to keep in view

Pause or stop new activity when needed. Existing positions and separately scheduled orders need their own review; stopping an agent does not automatically close them.

Costs and limits

A budget limits allocation, not losses. Review product fees, slippage and any leveraged exposure before activation.

View fees →

The next step stays yours.

Create an agent →

Start with a job you can describe

Decide whether you want a market-monitoring Autopilot, a specific rule-based strategy or a recurring purchase schedule. These are different workflows. Write down the budget, relevant assets, permitted actions and stopping condition before you activate anything.

The setup guide follows the path from draft to funding, authorization and the first execution check. The permissions guide explains why stopping new positions does not close existing exposure.

Review activity, not only the headline return

Inspect fills, fees, remaining allocation and skipped or failed actions. Automation can lose money and may need attention when market conditions, liquidity or your own objectives change. An illustrative strategy example is not a tested investment recommendation.

Continue with a related guide

Updated 2026-09-17. Sources, examples and review approach.

Create an agent →