You get a funded practice account of $10,000 and a week to use it. Real market data, real order books, simulated fills. Nothing you do can lose a cent, and nothing asks for a key.
At the end you get a table of which markets actually paid you and which quietly bled — then you decide whether to go near real money at all.
START THE PAPER WEEK ⬇ DOWNLOAD THE BOTIt parks a buy order and a sell order inside the gap between the best bid and the best ask. When both fill it has bought low and sold high on the same asset. It is not betting on direction.
It scans every market on the venue and refuses any that fails eight safety gates. The binding one: the spread must be at least 3× how far price moves in a minute. Most markets fail.
When nothing qualifies it holds no position and quotes nothing, and tells you why. Most days that is the correct answer. A bot that always trades is a bot with no standard.
Move the dials and watch the projection update against live market conditions. How wide you quote, how much per order, when to cut. It tells you when a setting is outside anything we have measured.
The bot trades your settings on a $10,000 paper account, rotating between markets as the edge moves. Every fill is recorded, including where the price went in the thirty seconds after.
A breakdown per market: fills, volume, profit, and profit per unit of volume. The last one is the number that matters, because it compares markets you spent six hours in against ones you spent six minutes in.
Lighter's rules ban "operating or controlling multiple accounts to earn additional points" and "trading between commonly controlled accounts", and they can reverse points after awarding them — before, during, or after the program.
So: one account per person, your own funds, your own key, real orders into the real book. Slower than the schemes going round, and it still stands when they get zeroed.