ClawLabs
All Projects
T

TraderBots

Beta

AI-powered trading bots that adapt, backtest, and execute autonomously.

TraderBots lets you describe a strategy in plain language and deploys AI-built bots across Hyperliquid, Binance, Bybit, GMX and Polymarket. Includes backtesting, risk management, and a regime-adaptive SuperBot.

Visit WebsiteView Source
AITradingFintechAutomationCrypto

Tech Stack

Python 3.12FastAPISQLAlchemyNext.js 14TypeScriptTailwind CSSshadcn/uiZustandCeleryRedisPostgreSQLInfluxDBDockernginxAnthropic ClaudeHyperliquid SDKPolymarket CLOBPytest

Key Features

  • Regime-adaptive SuperBot switching strategies automatically
  • Conversational AI bot builder powered by Claude
  • Event-driven backtesting with walk-forward validation
  • AI Scoring Engine (0–100 composite signal)
  • 3 automation modes: Alert, Semi-Auto, Autopilot
  • Multi-market support: Hyperliquid, Binance, Bybit, Polymarket

About this project

Overview

TraderBots is a full-stack AI trading bot platform that enables traders to build, backtest, and deploy automated strategies across multiple markets — Hyperliquid, Binance Futures, Bybit, GMX, and Polymarket — without writing a single line of code. At its core is the SuperBot, a regime-adaptive engine that switches between trend-following, mean-reversion, SMC, and news-driven strategies based on live market conditions.

How It Works

Users describe a strategy in plain language via a conversational AI builder powered by Claude (Anthropic). The system generates a full bot configuration including symbol selection, timeframe, entry/exit rules, and risk limits. Bots are then validated through an event-driven backtesting engine with realistic fees, slippage simulation, grid-search parameter optimization, and walk-forward out-of-sample validation. Once confident, users deploy in paper mode (zero risk, live prices) or go live in one of three automation modes: Alert, Semi-Auto, or Autopilot.

The AI Scoring Engine produces a composite 0–100 score from five weighted signals — technicals (30%), trend (20%), sentiment (15%), news (20%), and macro (15%) — to surface STRONG_LONG / LONG / SHORT / STRONG_SHORT signals in real time via WebSocket streams.

Use Cases

  • Retail traders automating crypto strategies without coding experience
  • Quant enthusiasts backtesting and optimizing SMC, EMA, Bollinger Band, or news-driven strategies
  • Copy trading marketplace — publish bots, earn revenue when others subscribe
  • Risk-managed autopilot trading with circuit breakers, daily loss limits, and a one-tap kill switch
  • Prediction market participants deploying event-driven bots on Polymarket

Frequently asked questions

Related projects

VideoPilot

Turn one idea into 100 viral videos with AI automation

ViralSync

Automate Reddit viral content into cross-platform social posts

EVBet.ai

AI-powered sports betting edge detector with +EV picks

Built with ClawLabs × Claude Code

This project was designed, built, and shipped with AI assistance. Every line of code crafted with Claude.

Learn more →