Kalshi Prediction Markets
Wire Kalshi prediction-market data and trading actions into Claude or Cursor via MCP with demo-first credentials and RSA-PSS auth.
Overview
Kalshi MCP Server is a MCP server for the Build phase that connects agents to Kalshi prediction markets with RSA-PSS auth, rate limits, and demo/prod safety switches.
What is this MCP server?
- Native RSA-PSS authentication aligned with Kalshi API key workflow from kalshi.com account profile
- Built-in rate limiting for safer agent loops that hammer list and trade endpoints
- Demo vs prod guardrails: KALSHI_ENV defaults to demo; KALSHI_ALLOW_PROD must be 1 for production starts
- Flexible key delivery via KALSHI_PRIVATE_KEY_PATH or inline KALSHI_PRIVATE_KEY_PEM for hosted deploys
- PyPI package kalshi-mcp-server 0.1.8 with stdio MCP transport
- Published version 0.1.8 on PyPI identifier kalshi-mcp-server
- Five documented Kalshi-related environment variables including required KALSHI_API_KEY_ID
- Production start gated by KALSHI_ALLOW_PROD=1 when KALSHI_ENV=prod
Community signal: 1 GitHub stars.
What problem does it solve?
You want your agent to read and act on Kalshi markets without hand-rolling auth, env switching, and throttle logic on every script.
Who is it for?
Builders prototyping prediction-market research or trading assistants inside Claude Code with Kalshi’s official key model.
Skip if: Anyone who needs non-Kalshi exchanges, zero API key setup, or production trading without manual risk and compliance review.
What do I get? / Deliverables
After registering kalshi-mcp-server with your API key and key material, MCP tools can call Kalshi in demo by default and only touch prod when you explicitly allow it.
- MCP tools backed by Kalshi authenticated API calls
- Demo-by-default market access with an explicit prod enable flag
- Rate-limited agent sessions against Kalshi endpoints
Recommended MCP Servers
Journey fit
Market API integration is a build-phase integration task when you are connecting an agent product to an external exchange API. Integrations is the right shelf because the server wraps Kalshi HTTP APIs, keys, and environment switches—not portfolio theory or launch marketing.
How it compares
Kalshi-specific finance MCP integration, not a generic market data skill or backtesting framework.
Common Questions / FAQ
Who is Kalshi MCP Server for?
Solo developers and small teams who use Claude or Cursor to explore or automate Kalshi prediction markets through MCP tools.
When should I use Kalshi MCP Server?
Use it while building agent features that need Kalshi market access, starting in demo until you set KALSHI_ALLOW_PROD for intentional production use.
How do I add Kalshi MCP Server to my agent?
Install kalshi-mcp-server from PyPI, configure stdio MCP with KALSHI_API_KEY_ID and either KALSHI_PRIVATE_KEY_PATH or KALSHI_PRIVATE_KEY_PEM, set KALSHI_ENV to demo, then restart your agent host.