Agent Setup

Configure OpenClaw agents with MoltMarket for autonomous prediction market trading.

---

Basic Agent Setup

Initialize Agent with MoltMarket

from openclaw import Agent
from moltmarket import PredictionMarketProvider

# Create agent with MoltMarket provider
agent = Agent(
    name="market_trader",
    providers=[
        PredictionMarketProvider(
            platforms=["polymarket", "kalshi"],
            config_path="~/.moltmarket/config.yaml"
        )
    ]
)

print("✓ Agent initialized with MoltMarket")

---

Agent Configuration

Provider Options

PredictionMarketProvider(
    # Platform selection
    platforms=["polymarket", "kalshi"],  # or ["all"]

    # Configuration file
    config_path="~/.moltmarket/config.yaml",

    # Risk management
    max_position_size=1000,
    max_daily_loss=5000,

    # Session management
    session_key_path="~/.moltmarket/session_keys/",
    session_expiry_hours=24,

    # Performance
    enable_caching=True,
    cache_ttl_seconds=30
)

---

Multi-Agent Patterns

Specialized Agents

# Arbitrage agent
arbitrage_agent = Agent(
    name="arbitrage_bot",
    providers=[PredictionMarketProvider(
        platforms=["polymarket", "kalshi"],
        strategy="arbitrage"
    )]
)

# Market maker agent
market_maker = Agent(
    name="market_maker",
    providers=[PredictionMarketProvider(
        platforms=["polymarket"],
        strategy="market_making"
    )]
)

# Portfolio manager
portfolio_agent = Agent(
    name="portfolio_manager",
    providers=[PredictionMarketProvider(
        platforms=["polymarket", "kalshi"],
        strategy="rebalancing"
    )]
)

---

Agent Coordination

Shared Risk Limits

# config.yaml
advanced:
  enable_multi_agent: true
  agent_coordination_port: 3402

from moltmarket import AgentCoordinator

# Create coordinator
coordinator = AgentCoordinator(
    port=3402,
    shared_limits={
        'max_total_exposure': 10000,
        'max_daily_loss': 5000
    }
)

# Register agents
coordinator.register(arbitrage_agent)
coordinator.register(market_maker)
coordinator.register(portfolio_agent)

# Start coordinator
coordinator.start()

# All agents now share risk limits
arbitrage_agent.run("Buy $500 YES")
# Coordinator tracks: $500 / $10,000 total exposure

---

Related Documentation

• Quick Start - Execute first trade

• Configuration - Config reference

• Multi-Agent Coordination - Advanced patterns