Prediction markets for
OpenClaw agents
Clawmarket enables AI agents to trade on Polymarket and Kalshi autonomously. Declarative market semantics. Non-custodial architecture. Production-ready.
from openclaw import Agent
from clawmarket import PredictionMarketProvider
# Enable prediction market trading
agent = Agent(
providers=[PredictionMarketProvider()]
)
# Agent trades on markets autonomously
agent.run("Buy $500 YES on Trump wins")Built for autonomous agents
Clawmarket eliminates the friction between AI agents and prediction markets through MCP-native tooling.
MCP-Native Protocol
Purpose-built Model Context Protocol server exposing prediction market operations as callable agent tools.
Declarative Semantics
Agents describe market intent using high-level parameters. Clawmarket handles API authentication, order execution, and settlement.
Non-Custodial Architecture
API keys and private keys remain under operator control. Configurable approval policies for every trade.
OpenClaw Native Provider
Drop-in integration for OpenClaw AI agents with native Python provider for autonomous market participation.
Cross-Platform Support
Unified interface for Polymarket (Polygon/USDC) and Kalshi (regulated US exchange). Single API for both platforms.
Real-Time Market Data
Orderbook depth, trade history, and market probabilities updated continuously. WebSocket feeds for live data.
Integrate in minutes
Transform natural language into prediction market trades. Drop-in integration for OpenClaw and custom agents.
// MCP Tool Call
{
"tool": "position.open",
"parameters": {
"market_id": "0x1a2b3c...",
"platform": "polymarket",
"outcome": "YES",
"amount": 100
}
}
// Response
{
"position_id": "pos_9x8y7z...",
"shares": 212.77,
"avg_price": 0.47
}Meet the Creator
Built by one person with a curiosity about AI and prediction markets. This is what happens when late-night questions turn into actual code.
Bryan Ang
Creator & Developer
Taiwan
Hi there! I'm Bryan, and I built Clawmarket because I kept wondering: what would happen if AI agents could actually participate in prediction markets? Not just analyze them, but trade on them, learn from them, make decisions.
I've spent years building software, but this project hits different. It sits right at the intersection of two things that genuinely fascinate me: AI that can reason about uncertainty, and markets that somehow turn chaos into collective wisdom. No fancy team, no investors breathing down my neck, just me, my laptop, and a lot of coffee.
The whole thing started when I realized there wasn't a good way for AI agents to securely trade on platforms like Polymarket and Kalshi. So I'm building it. Watching it come together, seeing the pieces click into place, that's the part I love.