DFlow is a comprehensive MCP server for accessing Kalshi prediction market data—the first CFTC-regulated exchange for trading on real-world events. It enables retrieval of real-time and historical market information through 23 specialized tools covering market analytics, trade history, and sentiment tracking. Key features of DFlow: - Event and market data retrieval with advanced search and filtering by tickers, mints, categories, and sports - OHLC candlestick time series and trade history for technical analysis and price action tracking - Forecast percentile history and analytics for monitoring market sentiment and predictions - Batch market queries and real-time data feeds supporting up to 100 simultaneous lookups - Access to market metadata, series information, and trade milestone data

DFlow MCP Server

[](https://smithery.ai/server/@openSVM/dflow-mcp)

Access Kalshi prediction market data - the first CFTC-regulated exchange for trading on real-world events. Built by OpenSVM. 23 tools for events, markets, trades, forecasts, candlesticks and live data.

Demo: https://dflow.opensvm.com

Features

This MCP server provides access to the complete Prediction Market Metadata API including:

Installation

Method 1: Install via Smithery (Recommended)

The easiest way to install this MCP server is through Smithery:

npx @smithery/cli install dflow-mcp-server --client claude

This will automatically configure the server for use with Claude Desktop.

Method 2: Manual Installation

#### Prerequisites

#### Install Dependencies

bun install

Usage

Starting the Server

# Development mode (with hot reload)
bun run dev
Production mode
bun start
Or directly with Bun
bun run src/index.ts

The server uses stdio transport for MCP communication, which is the standard for MCP clients.

Integration with MCP Clients

#### Claude Desktop

If you installed via Smithery, the configuration is automatic. For manual setup, add this to your Claude Desktop config:

{
  "mcpServers": {
    "dflow-mcp": {
      "command": "bun",
      "args": ["run", "/path/to/dflow-mcp/src/index.ts"]
    }
  }
}

#### Other MCP Clients

This server is compatible with any MCP client (Cursor, Continue, etc.). Use the same configuration format as above.

Available Tools

Event Tools

Market Tools

Trade Tools

Analytics Tools

Live Data Tools

Series Tools

Utility Tools

Example Tool Calls

Get a specific event

{
  "tool": "get_event",
  "arguments": {
    "event_id": "US-PRESIDENT-2024",
    "withNestedMarkets": true
  }
}

Get market by mint address

{
  "tool": "get_market_by_mint",
  "arguments": {
    "mint_address": "9WzDXwBbmkg8ZTbNMqUxvQRAyrZzDsGYdLVL9zYtAWWM"
  }
}

Get forecast history

{
  "tool": "get_forecast_percentile_history",
  "arguments": {
    "series_ticker": "US-PRESIDENT",
    "event_id": "US-PRESIDENT-2024",
    "percentiles": "25,50,75",
    "startTs": 1704067200,
    "endTs": 1706745600,
    "periodInterval": 3600
  }
}

Search events

{
  "tool": "search_events",
  "arguments": {
    "q": "election",
    "limit": 10,
    "sort": "volume",
    "order": "desc"
  }
}

Development

Project Structure

dflow-mcp/
├── src/
│   └── index.ts          # Main server implementation
├── package.json          # Dependencies and scripts
├── tsconfig.json         # TypeScript configuration
├── llms_dflow.json       # API specification
└── README.md             # This file

Building

bun run build

Testing

bun run test

API Details

The server implements the complete REST API defined in llms_dflow.json with the following base URL:

``` https://prediction-markets-api.dflow