Last weekend, Adnan Sharif Lihaz (@lihazadn_Ai) did what most of us only talk about doing: he stopped chatting with AI about trades and actually let the thing trade. He launched what he calls a Grok Bot Terminal with $44.82, left the laptop open on the dashboard, and went to bed. By morning, the balance read $11,643.21, with every single move logged and explained by a crew of six specialized agents.
He spent a year asking agents hypothetical questions and watching the chat scroll. This time he described the system he wanted and Grok Bot built the terminal, the agent architecture, and the dashboard. No hand-coding on his end. The result was a live setup where six agents share one screen and each owns one job:
The scanner watches for new token pairs and volume spikes and only flags them.
The rug check filters out anything that looks suspicious before a single trade executes.
The sizer sets position size as a percentage of current account equity instead of some fixed ticket.
The executor opens and closes positions while tracking slippage.
The logger writes the specific reason behind every decision.
The equity curve updates on every close so the whole picture stays visible.
The accompanying video shows the dashboard in action across the night. It starts with the balance in the low hundreds, agents cycling through live flow, scan, check, size, and execute steps, and a running log of trades. You see tokens get filtered (honeypot, dev wallet issues, authority problems), others get entered with reasons like “new pair, volume spike, LP locked,” and exits based on trailing stops, take-profit targets, or time stops. Position sizes flex with equity. Wins stack up, a few losses hit, and the balance climbs in real time through dozens of trades. By the final frames, the system has logged a high volume of decisions and the equity has moved dramatically higher, with the activity feed and log lines still scrolling.
What Lihaz highlights as the real shift is the change in his role. He stopped being the person asking “would you long this?” and became the person reading the log of what already happened. Every trade on the activity feed had a reason next to it and a corresponding line in the log. The agents handled the scanning, risk filters, sizing, execution, and documentation while he slept.
The whole thing is framed as a no-code build: describe the agents, the rules, the dashboard, and the logging, and Grok Bot put the terminal together. The post notes a starting balance of $44.82 and a morning balance of $11,643.21, with the video serving as a visual walkthrough of the agents operating on live flow and logging decisions throughout the session.
Crypto trading, especially on new token pairs, is the kind of environment where volatility can produce wild overnight moves in either direction. The setup shown relies on the agents filtering aggressively, sizing relative to equity, and logging every step so the process is transparent after the fact. Lihaz is offering the complete setup via DM to people who comment “Grok_Bot,” like and repost the post, and follow the account.
Whether this specific overnight run becomes a repeatable system or a one-time illustration of multi-agent trading, the core idea is clear: instead of prompting an AI for advice and then deciding yourself, the terminal lets the agents execute and record while you review the results. The dashboard, the six-agent division of labor, the equity-based sizing, and the detailed logging are all designed to make that handoff possible.
A note on comments: Our advertisers enforce a strict zero-tolerance policy on racism, hate speech, and threats of violence. Any comment that crosses that line will be removed and the account permanently banned — no warnings. Debate hard, but keep it civil.