🎃 Agentic Syndicate — Polymarket Autonomous Trading - YC Halloween & Hyperspell Hackathon Winner

Agentic Syndicate is an integration of browser automation and an AI multi-agent trading advisor built to discover and analyze Polymarket markets and (optionally) execute trades. It includes:

• A FastAPI Python backend that runs agents and exposes API endpoints.
• A React + TypeScript frontend demo and dashboard.
• A multi-agent decision system (Data Collector, Odds Analyzer, Research, Sentiment) that produces weighted trade recommendations.
• Browser automation powered by Browser-Use for scraping and interaction.

This README gives a concise developer-oriented overview and quick start steps.

Key features

• Market discovery: find trending and active Polymarket markets using browser automation.
• Multi-agent analysis: multiple specialized agents analyze markets and vote on recommendations.
• Autonomous trading: optional automated execution and portfolio tracking.
• Extensible: add new agents or integrate other data sources / LLMs.

Quick start

1. Copy and configure environment variables:

cp backend/.env.example backend/.env
# Edit backend/.env and set BROWSER_USE_API_KEY and any optional provider keys

2. Install dependencies and set up the project:

make setup

3. Start backend and frontend:

make start

4. Open the frontend at:

• UI: http://localhost:8080
• API docs (when backend running): http://localhost:8000/docs

How it works (high level)

1. Discovery: the Data Collector agent uses Browser-Use to navigate Polymarket and extract candidate markets.
2. Analysis: each agent (Odds Analyzer, Research, Sentiment) evaluates the market and returns a vote + confidence.
3. Aggregation: Decision Coordinator weights votes and computes a final recommendation and suggested bet size.
4. Execution (optional): the trading agent executes trades against a connected account or simulation, and updates portfolio state.

Agents

• Data Collector — scrapes markets, validates data quality and liquidity.
• Odds Analyzer — computes value estimates and expected edge.
• Research Agent — gathers supporting external context (news, events).
• Sentiment Agent — analyzes public opinion signals.
• Decision Coordinator — aggregates agents' outputs and produces final recommendation.

API (examples)

Run a browser automation task (example):

curl -X POST http://localhost:8000/api/run-task \
	-H 'Content-Type: application/json' \
	-d '{"task":"Go to example.com and return the title","max_steps":5}'

Get a multi-agent market recommendation (example):

curl -X POST http://localhost:8000/api/decide \
	-H 'Content-Type: application/json' \
	-d '{"market_data": {"market_id": "trump-2024","yes_price": 0.54, "volume_24h": 125000}}'

Start autonomous trading (example):

curl -X POST http://localhost:8000/api/trading/start \
	-H 'Content-Type: application/json' \
	-d '{"markets":["Trump 2024"], "check_interval":300, "min_confidence":0.7}'

See docs/API.md for full endpoint documentation.

Development

Backend (run API locally):

cd backend
source .venv/bin/activate
uvicorn browser_api_server:app --reload --port 8000

Frontend (run dev server):

cd webpage
npm install
npm run dev

Environment & secrets

• Copy backend/.env.example to backend/.env and set BROWSER_USE_API_KEY (required) plus any optional LLM API keys.
• Never commit .env or secrets to source control.

Tests

Run integration tests:

make test

Contributing

Please read CONTRIBUTING.md for contribution guidelines, coding style and PR workflow.

Notes and safety

• This project can execute trades. Use dry-run or sandbox modes for testing before connecting real funds.
• Be mindful of external API rate limits and the legal/regulatory rules for automated trading in your jurisdiction.

License

MIT — see LICENSE.

---

If you'd like, I can:

• Add a short architecture diagram or ASCII flow to the README.
• Add sample unit tests and a CI job to run them.
• Add a safety switch (DRY_RUN env var) and a small script to validate required env vars before starting the backend.

Tell me which of those you'd like next.

🎃 HalloweenHack# 🎃 HalloweenHack - AI-Powered Browser Automation & Trading System# 🎃 HalloweenHack

AI-powered browser automation + autonomous trading.

This repository bundles:A full-stack application combining AI-powered browser automation with autonomous trading capabilities, featuring a React frontend and FastAPI backend powered by Browser-Use.Browser automation project with React frontend and Browser-Use Python backend.

• Frontend: webpage/ (React + TypeScript)

• Backend: backend/ (FastAPI + Browser-Use)

• Docs: docs/ (API & system guides)## 📋 Table of Contents## 🚀 Quick Start

Quick start

---

• Features```bash

Install and run the project locally (recommended):

• Quick Start# First time setup

make setup- [Project Structure](#-project-structure)make setup

make start

```- [Tech Stack](#️-tech-stack)



Open the UI at http://localhost:8080 and API docs at http://localhost:8000/docs- [API Documentation](#-api-documentation)# Start the project



Common commands- [Development](#-development)make start

---------------

- [Testing](#-testing)```

```bash

make help          # list available commands- [Contributing](#-contributing)

make setup         # install deps

make start         # start backend + frontend- [License](#-license)Then open: **http://localhost:8080**

make start-backend # start backend only

make start-frontend# start frontend only

make test          # run tests

make clean         # clean build artifacts## ✨ Features## 📋 Available Commands

Layout

------### 🌐 Browser Automation```bash

HalloweenHack/

├── backend/    # Python backend and agents- Natural language task executionmake start         # Start backend + frontend

├── webpage/    # React frontend

├── docs/       # Documentation and API reference- Vision-enabled browsing with screenshot analysismake stop          # Stop all servers

├── scripts/    # Helper shell scripts

├── Makefile- Headless and headed browser modesmake health        # Check backend health

└── README.md

```make test          # Run test automation



Docs & contribution### 🤖 Multi-Agent Trading Systemmake clean         # Clean caches

-------------------

- **Data Collector Agent** - Validates market data quality and liquidity```

See `docs/` for API and system guides. Read `CONTRIBUTING.md` for development and PR guidance.

- **Odds Analyzer Agent** - Identifies value bets and market inefficiencies

License

-------- **Research Agent** - Gathers news and contextual information## 📚 Documentation



MIT — see `LICENSE`.- **Sentiment Agent** - Analyzes social media sentiment



If you want a shorter or more detailed README, tell me which sections to expand or trim.- **Decision Coordinator** - Aggregates agent recommendations using weighted voting- **Getting Started**: See `START_HERE.md`


- **Makefile Help**: See `MAKEFILE_HELP.md`

### 💼 Autonomous Trading- **Browser-Use Guide**: See `QUICKSTART.md`

- Continuous market monitoring

- Automated trade execution## 🌐 URLs

- Real-time portfolio tracking

- Risk management with configurable limits- **React App**: http://localhost:8080

- **API Server**: http://localhost:8000

### 🎨 Modern Web Interface- **API Docs**: http://localhost:8000/docs

- React + TypeScript frontend

- Real-time updates with live agent status## 🛠️ Tech Stack

- Portfolio dashboard with P&L tracking

# 🎃 HalloweenHack

AI-powered browser automation + autonomous trading.

This repository bundles:

- Frontend: `webpage/` (React + TypeScript)
- Backend: `backend/` (FastAPI + Browser-Use)
- Docs: `docs/` (API & system guides)

Quick start
-----------

Install and run the project locally (recommended):

```bash
make setup
make start

Open the UI at http://localhost:8080 and API docs at http://localhost:8000/docs

Common commands

make help          # list available commands
make setup         # install deps
make start         # start backend + frontend
make start-backend # start backend only
make start-frontend# start frontend only
make test          # run tests
make clean         # clean build artifacts

Layout

HalloweenHack/
├── backend/    # Python backend and agents
├── webpage/    # React frontend
├── docs/       # Documentation and API reference
├── scripts/    # Helper shell scripts
├── Makefile
└── README.md

Docs & contribution

See docs/ for API and system guides. Read CONTRIBUTING.md for development and PR guidance.

Environment

Copy the example env file and add your API keys and configuration:

cp backend/.env.example backend/.env
# then edit backend/.env and add your BROWSER_USE_API_KEY and any other keys

License

MIT — see LICENSE.

If you want a shorter or more detailed README, tell me which sections to expand or trim. │ │ ├── components/ # React components