KRA Racing Data Analysis MCP Server

What is KRA Racing Data Analysis MCP Server?

It is an MCP (Model Context Protocol) server that analyzes horse racing data from the Korea Racing Authority (KRA) public API. It provides tools, resources, and prompts for race analysis, horse performance evaluation, jockey statistics, odds analysis, and track condition assessment. It is designed for data analysts, developers, and racing enthusiasts.

How to use KRA Racing Data Analysis MCP Server?

Install Node.js 18+, clone the repository, run npm install, set the KRA_SERVICE_KEY environment variable (obtained from data.go.kr), then run in dev mode with npm run dev or build and start with npm run build and npm start. Register the server in your MCP client (e.g., Cursor) by adding the appropriate command and path in the mcp.json file.

Key features of KRA Racing Data Analysis MCP Server

• Analyze races by date, race number, and track.
• Evaluate horse performance and form.
• Retrieve jockey statistics and performance.
• Analyze odds, patterns, and market trends.
• Assess track condition and weather impact.
• Access race schedules, horse details, and track info via resources.
• Generate race predictions and performance reports via prompts.

Use cases of KRA Racing Data Analysis MCP Server

• Predict race outcomes using data‑driven analysis.
• Build betting strategies with odds and volume patterns.
• Evaluate horse, jockey, and trainer performance over time.
• Analyze track and weather effects on race results.
• Manage investment risk through probability models.

FAQ from KRA Racing Data Analysis MCP Server

What are the runtime requirements?

Node.js 18.0.0 or higher, npm or yarn, and a KRA public API service key.

How do I get the KRA API key?

Apply for the “한국마사회 경주기록 정보” API on the Korean Public Data Portal (data.go.kr), then set the approved service key as the KRA_SERVICE_KEY environment variable or in a .env file.

Does this server use live KRA data?

Currently it uses mock data; integration with the official KRA API is planned for a future release.

How do I register this server in Cursor IDE?

Add a JSON entry to ~/.cursor/mcp.json with the command npx tsx (dev) or node (prod) pointing to the server’s entry file.

What analysis tools are available?

analyze-race, analyze-horse-performance, get-jockey-stats, analyze-odds, and analyze-track-condition.