TradeClaw

<p align="center">
  <img src="assets/images/tradeclaw.png" alt="TradeClaw" width="480" />
</p>

<h1 align="center">TradeClaw — LLM Agent Trading System</h1>

<p align="center">
  <a href="license.txt"><img src="https://img.shields.io/badge/license-MIT-blue.svg" alt="License" /></a>
</p>

<p align="center">AI-powered autonomous trading system for US equities and ETFs.</p>

<p align="center">🇨🇳 <a href="README_CN.md">中文文档</a></p>

## Highlights

- **LLM-Driven Decisions** — LangGraph ReAct Agent with zero hard-coded rules
- **Event-Driven Architecture** — Async event queue, fully decoupled components
- **Multiple Workflows** — Sequential, tool-calling, Black-Litterman, cognitive arbitrage, and more
- **Flexible LLM Config** — Multi-provider / multi-model, per-agent override, YAML-persisted
- **Real-Time Monitoring** — WebSocket quotes + news polling (AkShare / Alpaca / Tiingo / Finnhub) with LLM importance scoring
- **Configurable Risk Rules** — Stop-loss / take-profit rule chains (YAML); hard rules and LLM-triggered analysis coexist
- **Browser Automation** — Playwright-driven dynamic web scraping and interaction
- **Code Execution Sandbox** — RestrictedPython sandbox (local) or OpenSandbox (Docker isolation)
- **Telegram Control** — Remote monitoring and command execution
- **Modern Web UI** — React + TypeScript + TailwindCSS responsive dashboard

## Quick Start (Docker — Recommended)

The fastest way to get started is the **one-line install script**:

```bash
curl -fsSL https://raw.githubusercontent.com/hugging-leg/TradeClaw/main/install.sh | bash
```

This will:

1. Create a `tradeclaw/` directory
2. Download `docker-compose.yml`, `env.template`, and SearXNG config
3. Create the `user_data/` directory tree
4. Copy `env.template` → `.env` for you to edit
5. Start all services via `docker compose up -d`

After the script finishes, edit `tradeclaw/.env` with your API keys, then visit **http://localhost:8000**.

### Manual Docker Setup

```bash
git clone https://gi