Kalshi MCP — Prediction Markets Server¶
What It Is¶
Kalshi is the first CFTC-regulated prediction market in the US. This MCP server gives AI agents programmatic access to Kalshi markets — enabling programmatic event-driven analysis and trading.
GitHub: https://github.com/onofre-jauregui/kalshi-mcp (TypeScript, new)
Tools Available¶
| Tool | Description |
|---|---|
get_markets |
List available prediction markets with prices and volume |
get_positions |
Check current positions across all markets |
place_order |
Place buy/sell orders on event contracts |
get_market_history |
Historical pricing and settlement data |
Quick Start¶
# Clone and install
git clone https://github.com/onofre-jauregui/kalshi-mcp.git
cd kalshi-mcp
npm install
# Configure environment
export KALSHI_API_KEY="your-kalshi-api-key"
export KALSHI_PRIVATE_KEY="your-private-key-path"
# Add to Hermes
hermes mcp add kalshi --command "node" --args "dist/index.js" --workdir "$(pwd)" \
--env KALSHI_API_KEY KALSHI_PRIVATE_KEY
Manual Configuration¶
{
"mcpServers": {
"kalshi": {
"command": "node",
"args": ["dist/index.js"],
"env": {
"KALSHI_API_KEY": "your-kalshi-api-key",
"KALSHI_PRIVATE_KEY": "/path/to/private-key.pem"
}
}
}
}
Prerequisites¶
- Kalshi Account: Sign up at kalshi.com
- API Key: Generate from Kalshi developer dashboard
- RSA Private Key: Required for authenticated API calls (Kalshi uses RSA-based auth)
Business Use Cases¶
- Event-Driven Analysis: Have AI agents monitor prediction market probabilities for real-time event forecasting
- Market Scanning: Scan all active markets for mispriced event contracts
- Portfolio Monitoring: Track event-contract portfolio performance and risk exposure
- Programmatic Hedging: Use prediction markets to hedge against defined-event risks
Limitations¶
- US-only (Kalshi is CFTC-regulated for US residents)
- Kalshi API requires RSA key-pair authentication (more complex than Bearer tokens)
- New server — limited tool coverage, expect rapid iteration
- Prediction markets are not available in all US states
See Also¶
- SentiSense MCP — for market sentiment intelligence
- HPSILab Quant Finance MCP — for options analytics and Monte Carlo
- pipeworx-io/mcp-tradier — for traditional stock/options market data