Your capital is ready to deploy
Funds are confirmed and sitting in cash. Choose a strategy mandate and activate the swarm, your agents start scanning markets and opening positions within seconds.
Equity curve
Allocation by category
Open positions
| Market | Side | Entry | Now | Size | P&L | Agent |
|---|
Your agents
Risk & quality
Swarm decisions
Live activity
Live prediction markets
Account statement
Refer & earn
Each funded referral drops your performance fee by 1% (floor 12%) for as long as they stay invested.
Referral leaderboard
| # | Allocator | Funded referrals | Earned |
|---|
Transaction history
| Type | Amount | Asset | Detail | Date | Status |
|---|
Account
Preferences
Security
How ORACLE works
A team of Claude Code agents trades prediction markets on your behalf. You stay in control of the capital, they do the research and execution. Here's the whole loop, start to finish.
Fund your account
Deposit from $1,000 in USDC or USDT on any supported network. Funds land in segregated cold-storage custody and show up as idle cash on your dashboard within a few network confirmations.
Choose your strategy profile
Choose your strategy profile: Conservative, Balanced, Aggressive, or Momentum. Your potential profit depends on the profile you select. This strategy determines your mandate — including the number of agents involved, their level of influence, and the market categories they will focus on. You can change it anytime.
Agents go to work
Each Claude Code agent owns a category on Polymarket. It reads polls, news, on-chain flow and the order book, estimates the true probability of an outcome, and takes the side it believes is mispriced, logging a written rationale for every position. You watch it all live in Portfolio and Markets.
Keep the profit, withdraw anytime
You'll keep 80% of your net profit; Oracle takes 20%, which only exceeds your larger profit. There's no lock: withdraw your available balance to any wallet instantly.
Common questions
Four ways to run the swarm
A mandate is the rulebook your agents trade under. It sets how many agents deploy, how much leverage they may use, which categories they focus on, and the drawdown ceiling that pauses trading. Switch any time, the swarm rebalances into the new mandate within minutes. Targets are net of the 20% performance fee and are objectives, not guarantees.
How the swarm finds edge
Every position is produced by the same controlled pipeline inside the client workspace: data ingestion, probability modeling, risk checks, execution and post-trade review.
Market intake
Agents scan prediction markets, order-book flow, news, filings, polls, on-chain signals and pricing deltas across categories.
Fair-value model
Each candidate is repriced against source evidence. Only markets with measurable spread between model probability and live odds move forward.
Risk gate
Position size, leverage, category concentration and drawdown limits are checked before capital is committed.
Execution & review
Approved trades are logged with rationale, monitored until resolution and folded back into the next allocation cycle.