Polytrader AI Project

Overview

Welcome to the Polytrader AI project! This repository consists of two main parts:

• backend/: A Python-based AI agent using LangChain, LangGraph, and related libraries to analyze and trade on Polymarket.
• frontend/: A Next.js application with TailwindCSS, Shadcn UI, and TypeScript for the user-facing interface.

Poly is the trading agent, and it is capable of:

• Performing deep research on a given market using Exa and a recursive process to ask better and more refined questions about the market in question.
  • Inspiration for my implementation of Deep Research comes from dzhng/deep-research, which can be found here.
• Analyze the market conditions, and produce a summary of the current state of the market.
• Decide whether to buy, sell, or hold a position, based on the market conditions and the agent's own analysis.

On the web app, you can see the agent's thought process and, if it does decide to make a trade, you can confirm or reject the trade.

Getting Started (High-Level)

1. Clone this repository

   git clone https://github.com/YourUser/polytrader.git
   cd polytrader

2. Set up Environment Variables

Copy .env.example in the backend/ directory (and similarly .env.example in frontend/ if available) to .env (or .env.local for the frontend), then fill in the required values. For the backend, ensure the environment variables that your agent or Graph nodes need (like API keys) are properly populated. For the frontend, any variables you need at build time can be placed in .env.local or shared with NEXT_PUBLIC_ prefix. Backend Setup See backend/README.md for details on installing Python dependencies, running the AI agent, or using the LangGraph server.

3. Frontend Setup See frontend/README.md for more information about installing Node.js dependencies and starting the Next.js dev server.

4. Running Locally

   • Backend: Typically run make lg-server from the backend/ directory to start the agent’s local dev environment.
   • Frontend: From the frontend/ folder, install dependencies with pnpm install and run pnpm dev.

What you will need

• OpenAI API key
• Exa API key
• Tavily (not actively being used) API key
• Web3 wallet (e.g. MetaMask, Phantom, etc.)

Configuring Polymarket

A detailed guide on how to configure Polymarket can be found in the backend/README.md file, here.

Structure

├── backend/
│ ├── src/
│ ├── tests/
│ ├── Makefile
│ ├── ...
│ └── pyproject.toml
├── frontend/
│ ├── app/
│ ├── components/
│ ├── lib/
│ ├── types/
│ ├── ...
│ ├── package.json
│ └── pnpm-lock.yaml
└── README.md (this file)