Building a Football Fixture MCP Server with Sportmicro API

Описание и применение

Небольшой сервер MCP на TypeScript, предоставляющий пять инструментов для исследования футбольных матчей через Sportmicro API. Валидация входов, структурированные ответы и разделение логики протокола и клиента делают интеграцию надёжной и простотой в поддержке.

Задача

Expose clean, validated football fixture data to AI clients without a messy integration of loosely validated endpoints.

Результат

A small, maintainable MCP server that exposes five focused tools for football fixture research, validates inputs early, and returns structured JSON responses, enabling reliable AI client interactions.

Рабочий процесс

• Define MCP tools and input schemas in server.ts
• Implement Sportmicro client with typed HTTP requests in sportmicro-client.ts
• Validate tool input with zod before making API calls
• Send requests to Sportmicro using Bearer token
• Return structured JSON payloads or error payloads to the MCP client
• Test validation and error handling with Node test runner
• Deploy server via node dist/index.js

Доказательства

I built this project around a simple problem: football fixture data is easy to query in isolation, but hard to expose cleanly to AI clients

Источник

The result is a TypeScript MCP server built with @modelcontextprotocol/sdk, zod, and a typed Sportmicro client.

Источник

The server is intentionally read-only. It does not mutate any Sportmicro data, and it only exposes the documented football endpoints needed for fixture research.

Источник