🔮 OMEN — The Oracle Machine

Thousands of AI minds debate. One verdict. You profit.

🌐 Live Demo • Quick Start • Features • How It Works • Pricing • Trading • API Docs

---

🌐 Live Demo

➡️ Try OMEN Live

Full access to the AI Oracle, Auto-Pilot, Backtesting, Whale Tracking, and more.

🎯 What is OMEN?

OMEN is an AI-powered prediction and copy-trading platform for Polymarket. It combines:

• 🧠 Swarm Intelligence — 45 AI agents debate outcomes and reach consensus
• 🐋 Whale Intelligence — Live on-chain tracking of top Polymarket wallets on Polygon
• 📊 Real Trading — Manual, Oracle-driven, and copy-trade execution via Polymarket CLOB
• 💰 Crypto Payments — Pay with MATIC/USDC on Polygon. No subscriptions.

Users don't configure APIs. They don't set up models. They don't study charts. They ask OMEN a question. The swarm deliberates. A verdict appears. Then one button: "Bet with the Oracle."

---

📸 Screenshots

Dashboard — System overview with quick Oracle access

🔮 Oracle War Room	🐋 Whale Tracker
45 AI agents debate live with research-driven swarm votes	Live on-chain whale tracking on Polygon
📊 Trading	💳 Credits
Manual, Oracle auto-trade, and whale copy-trade	Crypto payments on Polygon with tiered credit packages
🏆 Leaderboard	⚡ God View
Live Polymarket top 100 traders with PnL tracking	45-agent dark terminal with research-informed multi-round debate
🤖 Auto-Pilot	📈 Portfolio
Automated trading with Conservative/Balanced/Aggressive risk profiles	Track positions, P&L, and trade history
🔬 Backtest
Backtest Oracle predictions against real Polymarket outcomes

---

✨ Features

🔮 Oracle Chamber — Real AI Debates

The Oracle uses a 5-agent swarm powered by Gemini 2.0 Flash with distinct personalities:

Agent	Role	Style
🟢 Atlas	Bull Analyst	Finds reasons to buy
🔴 Nemesis	Bear Analyst	Finds reasons to sell
🔵 Quant	Statistician	Pure data and probability
🟡 Maverick	Contrarian	Challenges consensus
🟣 Clio	Historian	Historical patterns and precedent

Agents make real LLM API calls → Debate independently → Vote → Swarm matrix animates → Verdict with confidence score.

⚔️ War Room — 45 AI Agents Vote Live

• Each agent card reveals sequentially with real AI reasoning
• 45-agent swarm network with ripple animation shows consensus forming
• Green = YES, Red = NO — watch the swarm converge in real-time
• Whale agreement dots overlay for confidence boost
• Final verdict card with confidence percentage

🐋 Whale Intelligence — Live On-Chain

• Live Polygon blockchain tracking via polygon-bor-rpc.publicnode.com
• Real MATIC and USDC balances for tracked wallets
• Transaction count and recent volume monitoring
• Polymarket CTF Exchange detection — flags trades to Polymarket contracts
• DB whales with historical stats: win rate, PnL, volume, specialty
• Copy any whale — one click to mirror their latest trade

📊 Trading — Three Modes

Mode	Description
Manual	Browse live Polymarket markets → select → set price/size → place limit order
Oracle Auto-Trade	Ask Oracle → AI analyzes → auto-places bet on matching Polymarket market
Copy Trade	Enter whale address → mirrors their latest Polymarket position

Safety built-in:

• 🔒 API-only credentials — no private keys stored, only Polymarket API creds
• 📊 Liquidity check — rejects trades on illiquid orderbooks (spread > $0.50)
• 🛡️ Risk controls — max bet $50, daily limit $200, stop-loss 35%
• 🔐 Fernet encryption — all credentials encrypted at rest

💰 Crypto Payments — Polygon

• Wallet: 0x135C480C813451eF443A2F60cfaD49EA7197B855
• Accepted: MATIC, USDC, USDC.e on Polygon
• Verification: On-chain transaction hash verification via Polygon RPC
• Auto-credit: Credits added automatically after blockchain confirmation

---

🏗️ How It Works

┌─ USER ──────────────────────────────────────────────────┐
│  "Will Bitcoin exceed $150k by end of 2026?"            │
└──────────────────────┬──────────────────────────────────┘
                       │
┌──────────────────────▼──────────────────────────────────┐
│                 🔮 ORACLE ENGINE                        │
│           Gemini 2.0 Flash via OpenRouter               │
│                                                         │
│  ┌─────────┐ ┌─────────┐ ┌─────────┐ ┌─────┐ ┌─────┐  │
│  │  Atlas   │ │ Nemesis │ │  Quant  │ │ Mav │ │Clio │  │
│  │  (Bull)  │ │ (Bear)  │ │ (Stats) │ │     │ │     │  │
│  └────┬─────┘ └────┬────┘ └────┬────┘ └──┬──┘ └──┬──┘  │
│       └────────────┼──────────┼─────────┼───────┘      │
│                    ▼          ▼         ▼              │
│         ╔════════════════════════════════════╗          │
│         ║  45 SWARM AGENTS VOTE           ║          │
│         ║  663 YES / 537 NO                  ║          │
│         ╚════════════════════════════════════╝          │
│                         │                              │
│              ╔══════════▼════════════════╗              │
│              ║  VERDICT: YES — 72%       ║              │
│              ╚═══════════════════════════╝              │
└──────────────────────┬──────────────────────────────────┘
                       │
┌──────────────────────▼──────────────────────────────────┐
│                 🐋 WHALE LAYER (Polygon)                │
│  Live blockchain tracking via RPC                       │
│  Whale agreement: 5/8 → Confidence BOOST                │
└──────────────────────┬──────────────────────────────────┘
                       │
┌──────────────────────▼──────────────────────────────────┐
│                 📊 TRADING ENGINE                        │
│  Manual | Oracle Auto-Trade | Copy Trade                │
│  User's Polymarket API → py-clob-client → CLOB          │
└──────────────────────┬──────────────────────────────────┘
                       │
┌──────────────────────▼──────────────────────────────────┐
│                 💰 PAYMENT LAYER                         │
│  Polygon blockchain: MATIC/USDC → On-chain verify       │
└─────────────────────────────────────────────────────────┘

---

💰 Pricing

Getting Started — 50 Free Credits

Every new account gets 50 free credits to try everything.

Credit Packages (Crypto — Polygon)

Package	Credits	Price	Rate
Starter	50 credits	$5	10 credits/$1
⭐ Popular	120 credits	$10	12 credits/$1
Pro	300 credits	$20	15 credits/$1
🐋 Whale	1,000 credits	$50	20 credits/$1

What Credits Cost

Action	Cost
🔮 Oracle Prediction	1 credit
🐋 Whale Deep Dive	1 credit
🤖 Auto-Trade Bet	2 credits
🏆 Leaderboard	Free

---

🚀 Quick Start

Quick Deploy (Recommended)

git clone https://github.com/Mecasa-hub/omen.git
cd omen
cp .env.example .env  # Add your OpenRouter API key
pip install aiosqlite httpx aiohttp pydantic uvicorn fastapi cryptography py-clob-client
python -m uvicorn deploy:app --host 0.0.0.0 --port 8888
# OMEN is live at http://localhost:8888

Docker

git clone https://github.com/Mecasa-hub/omen.git
cd omen
cp .env.example .env
docker-compose up -d

Environment Variables

# Required
OPENROUTER_API_KEY=your-openrouter-key
LLM_MODEL=google/gemini-2.0-flash-001

# Payment wallet (pre-configured)
PAYMENT_WALLET=0x135C480C813451eF443A2F60cfaD49EA7197B855

# Polymarket (users provide their own API keys in the UI)
# No server-side trading keys needed

---

📁 Project Structure

omen/
├── deploy.py           # FastAPI app — all routes + Oracle engine (31KB)
├── ui.html             # MiroFish-inspired frontend (37KB)
├── trading.py          # Polymarket CLOB trading module (14KB)
├── whale_tracker.py    # Live Polygon blockchain whale tracking
├── payments.py         # Polygon crypto payment verification
├── .env                # API keys and configuration
├── data/               # SQLite database
├── docs/
│   ├── screenshots/    # All 10 pages: Dashboard, Oracle, Whales, Trading, Credits, Leaderboard, God View, Auto-Pilot, Portfolio, Backtest
│   ├── API.md          # Full API reference
│   ├── ARCHITECTURE.md # System design
│   ├── DEPLOYMENT.md   # Production setup guide
│   ├── CREDITS.md      # Credit system mechanics
│   └── VIRAL_STRATEGY.md
├── backend/            # Original modular backend (reference)
├── frontend/           # Vue 3 source (optional)
├── tests/              # Test suite
├── scripts/            # DB migration, whale seeding
├── docker-compose.yml
└── README.md

---

🛠️ Tech Stack

Layer	Technology
Backend	FastAPI, Python 3.11+, aiosqlite
Frontend	Embedded HTML/CSS/JS (MiroFish aesthetic)
AI Engine	Gemini 2.0 Flash via OpenRouter (multi-agent swarm)
Trading	py-clob-client → Polymarket CLOB API
Blockchain	Polygon RPC (whale tracking + payment verification)
Payments	MATIC/USDC on Polygon (on-chain verification)
Security	JWT auth, Fernet-encrypted credentials
Fonts	JetBrains Mono + Space Grotesk

---

🗺️ Roadmap

🧠 Phase 3: Intelligence Suite

Feature	Description
Advanced Swarm	45 AI agents across 9 categories (Technical, Macro, Sentiment, Crypto, Sports, Risk, Timing, Fundamental)
Auto-Pilot	Fully automated trading with 3 risk profiles (Conservative, Balanced, Aggressive)
Backtesting	Test Oracle accuracy against real resolved Polymarket markets
Portfolio Tracker	Track positions, PnL, win rate, trade breakdown
Alert System	Real-time notifications for whale moves, Oracle streaks, trade results
Whale Discovery	Auto-detect profitable wallets from Polymarket trade data

✅ Phase 1 — Foundation (Complete)

• 🔮 Oracle Engine with 5-agent AI swarm (Gemini 2.0 Flash)
• 🎨 MiroFish-inspired minimal UI (black & white theme)
• ⚔️ War Room with 45-agent animated swarm network
• 🐋 Live whale tracking on Polygon blockchain
• 💰 Crypto payments (MATIC/USDC on Polygon)
• 📊 Polymarket market discovery via Gamma API
• 🔐 JWT authentication + encrypted credential storage
• 📈 Dashboard with system stats

✅ Phase 2 — Trading (Complete)

• 📊 Manual trading via Polymarket CLOB
• 🔮 Oracle auto-trade (AI verdict → auto-bet)
• 🐋 Copy trading (mirror whale positions)
• 🛡️ Liquidity checks + risk controls
• 🔐 Per-user encrypted API credential storage
• 📜 Trade history logging

✅ Phase 3 — Intelligence (Complete)

• 🧬 Advanced swarm — 50+ agent personas with specialized strategies
• 📊 Backtesting engine — test strategies against historical data
• 🐋 Whale discovery — auto-detect new profitable wallets
• 📈 Portfolio tracker — aggregate positions, PnL, win rate
• 🔔 Alert system — whale moves, Oracle streaks, market events

✅ Phase 4 — Auto-Pilot (Complete)

• 🤖 Auto-pilot mode — fully automated trading based on Oracle + whales
• 📱 Mobile app (React Native)
• 🌐 Multi-chain support (Azuro, Overtime Markets)
• 💬 AI chat assistant with memory per user
• 🤖 Telegram bot for alerts and quick trades
• 🎬 Brag cards + social sharing for viral growth

🌍 Phase 5 — Ecosystem

• 🏪 Whale marketplace — subscribe to top traders' strategies
• 🏆 Public leaderboard with SEO optimization
• 🤝 Referral system — 10% credit bonus for invites
• 🐦 X/Twitter whale alert bot
• 🌍 Multi-language support
• 🏛️ DAO governance for platform decisions

---

📚 Documentation

Doc	Description
Architecture	System design & data flow
API Reference	All endpoints with examples
Deployment Guide	Production setup
Credit System	Pay-as-you-go mechanics
Viral Strategy	Growth hacking playbook

---

📄 License

MIT License — see LICENSE for details.

---

🔮 The swarm has spoken. Will you listen?

⭐ Star this repo • 🐛 Report Bug • 💡 Request Feature

⚡ God View Terminal (MiroFish-Inspired)

A dark hacker-style terminal interface inspired by the MiroFish God View for real-time swarm intelligence visualization.

Features

• 45 Named AI Agents — Each with unique personality, backstory, cognitive biases, and expertise
• Research-Informed Predictions — Agents receive real-time research context via Gemini 2.0 Flash before voting
• Multi-Round Debate System:
  • Round 1: All 45 agents vote independently with research context
  • Round 2: Top 15 most confident agents re-evaluate after seeing Round 1 results
  • Faction Formation: Agents cluster into Strong Bulls, Cautious Bulls, Strong Bears, Cautious Bears
  • Vote Flips: Track which agents changed their minds during debate and why
• Real-Time Event Stream — Scrolling terminal log of agent activities
• Performance Metrics — Throughput, latency, OPS/s, memory, CPU, queue, errors, tokens
• D3.js Force Graph — 45 agent nodes with connections and operation labels
• Boot Animation — Agents come online 0/45 → 45/45 with status messages
• Obsidian-Style Memory Vault — Each agent has persistent memory stored in markdown files
• Autoresearch-Inspired Learning — Agents track hypotheses, evidence, and post-mortems

Agent Personalities (45 Unique Agents)

Agent	Role	Personality	Risk
ORCHESTRATOR	Master Controller	Neutral coordinator	5/10
SENTINEL	Risk Scanner	Cautious, threat-focused	2/10
VECTOR	Data Embedder	Pure quantitative	5/10
CIPHER	Encryption/Privacy	Paranoid analyst	3/10
BEACON	Trend Discoverer	Optimistic spotter	7/10
FLUX	Momentum Trader	Aggressive trend follower	8/10
FROST	Cold Logic	Zero emotion, pure probability	4/10
BOLT	Speed Trader	First to react	9/10
THORN	Contrarian	Bets against the crowd	7/10
RAVEN	Intelligence Scout	Obscure data sources	6/10
...and 35 more	Each uniquely crafted

API Endpoints

• GET /api/godview/agents — List all 45 agents with full profiles
• POST /api/godview/predict — Run full research-informed multi-round prediction
• GET /api/godview/stats — Real-time system metrics
• GET /api/godview/agent/{id}/stats — Individual agent stats and vote history

Prediction Response

{
  "question": "Will Bitcoin exceed $150k by end of 2026?",
  "votes": [...],
  "consensus": {"direction": "NO", "percentage": "37.8", "yes_count": 17, "no_count": 28},
  "research_context": "## Research Brief: Bitcoin $150k...",
  "rounds": [{"round": 1, "name": "Initial Vote", "yes": 17, "no": 28}, {"round": 2, "name": "Debate", "yes": 17, "no": 28, "changes": 4}],
  "factions": [{"name": "Strong Bulls", "count": 9, "leader": "FLUX"}, ...],
  "vote_changes": [{"codename": "DELTA", "from": "NO", "to": "YES", "reason": "..."}, ...]
}