Sports MCP Server

Sports MCP Server

A unified MCP server that combines The Odds API and SportMonks Football API v3 to provide 28 tools for sports data like odds, scores, fixtures, standings, and predictions, enabling Claude to interact with a wide range of sports information.

Category
访问服务器

README

🏟️ Sports MCP Server

A Model Context Protocol (MCP) server that unifies The Odds API and SportMonks Football API v3 into a single tool surface for Claude Desktop and Claude Code.

What This Does

Source Coverage Tools
The Odds API All sports — NBA, NFL, NHL, NCAAM, NCAAF, Soccer, Tennis, MMA, Golf, MLB 10 tools: odds, scores, events, markets, participants, historical
SportMonks Football (soccer) deep data — EPL, La Liga, Bundesliga, Ligue 1, Serie A, Champions League, etc. 18 tools: fixtures, standings, statistics, xG, predictions, odds, livescores, squads

Total: 28 tools available to Claude via MCP.


Quick Start

1. Install dependencies

cd sports-mcp-server
npm install
npm run build

2. Set your API keys

You'll configure these in the Claude Desktop config (step 3), but for testing:

export ODDS_API_KEY="your-odds-api-key-here"
export SPORTMONKS_TOKEN="your-sportmonks-token-here"

3. Configure Claude Desktop

Add this to your Claude Desktop config file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "sports-data": {
      "command": "node",
      "args": ["/FULL/PATH/TO/sports-mcp-server/dist/index.js"],
      "env": {
        "ODDS_API_KEY": "your-odds-api-key-here",
        "SPORTMONKS_TOKEN": "your-sportmonks-token-here"
      }
    }
  }
}

⚠️ Replace /FULL/PATH/TO/ with the actual absolute path to the project.

4. Restart Claude Desktop

After saving the config, restart Claude Desktop. You should see the sports tools available in the MCP tools menu.


For Claude.ai (Web / Chrome) — ⭐ This is what you want

The web app needs an HTTP server, not stdio. Here's how:

Step 1: Build & start the HTTP server

cd sports-mcp-server
npm install
npm run build

# Start the HTTP server with your API keys
ODDS_API_KEY=your-key-here SPORTMONKS_TOKEN=your-token-here npm run start:http

You should see:

🏟️  Sports MCP Server (HTTP/SSE) running on http://localhost:3456
   Tools: 10 Odds API + 18 SportMonks = 28 total
   Odds API key:     ✅ configured
   SportMonks token: ✅ configured

Step 2: Connect in Claude.ai

  1. Go to claude.ai → Settings (bottom-left gear icon)
  2. Click Integrations in the left sidebar
  3. Click Add Custom Integration
  4. Set:
    • Name: Sports Data
    • URL: http://localhost:3456/sse
  5. Save and enable it

That's it — the 28 tools are now available in your Claude.ai conversations.

Optional: Protect with auth token

ODDS_API_KEY=xxx SPORTMONKS_TOKEN=xxx MCP_AUTH_TOKEN=my-secret npm run start:http

Then add the token in the integration config as a Bearer token.

Optional: Run on a remote server

If you want it always-on (e.g. on a VPS or Render/Railway):

  1. Deploy the project to your server
  2. Run npm run start:http with your env vars
  3. Use https://your-server.com/sse as the integration URL
  4. Set MCP_AUTH_TOKEN for security

Keep it running (local)

Use pm2 or a simple background process:

# Install pm2 globally
npm install -g pm2

# Start as a daemon
ODDS_API_KEY=xxx SPORTMONKS_TOKEN=xxx pm2 start dist/server-http.js --name sports-mcp

# Auto-restart on reboot
pm2 startup
pm2 save

For Claude Desktop (local app)

See the config under Quick Start above — uses stdio transport via dist/index.js.


For Claude Code

claude mcp add sports-data node /FULL/PATH/TO/sports-mcp-server/dist/index.js \
  -e ODDS_API_KEY=your-key \
  -e SPORTMONKS_TOKEN=your-token

Tool Reference

The Odds API Tools (prefix: odds_)

Tool Description Quota Cost
odds_get_sports List all available sports and their keys Free
odds_get_events List upcoming/live events for a sport Free
odds_get_odds Primary tool — get odds for upcoming/live games (ML, spreads, totals) markets × regions
odds_get_scores Live scores + recently completed games 1-2
odds_get_event_odds Odds for a single event — supports ALL markets including player props unique markets × regions
odds_get_event_markets Discover available market keys for an event 1
odds_get_participants List teams/players for a sport 1
odds_get_historical_odds Historical odds snapshot at a timestamp 10 × markets × regions
odds_get_historical_events Historical events at a timestamp 1
odds_get_historical_event_odds Historical odds for single event — all markets 10 × markets × regions

Common sport keys: basketball_nba, basketball_ncaab, americanfootball_nfl, americanfootball_ncaaf, icehockey_nhl, soccer_epl, soccer_germany_bundesliga, soccer_spain_la_liga, soccer_france_ligue_one, tennis_atp_*, tennis_wta_*

SportMonks Tools (prefix: sm_)

Tool Description
sm_livescores Live soccer scores (inplay / all today / latest updated)
sm_fixtures Fixtures by ID, date, date range, team, H2H, or search
sm_leagues Leagues — all, by ID, live, by country, by team, search
sm_seasons Seasons — all, by ID, by team, search
sm_standings League tables — by season, round, or live
sm_teams Teams — all, by ID, country, season, search
sm_players Players — by ID, country, search
sm_team_squads Team squad (current or by season)
sm_statistics Season/stage/round statistics for teams or players
sm_topscorers Top scorers by season or stage
sm_predictions Probabilities and value bets (SM's own model)
sm_expected_xg Expected goals (xG) by team or player
sm_odds_prematch Pre-match odds by fixture, bookmaker, or market
sm_odds_inplay In-play odds by fixture, bookmaker, or market
sm_schedules Match schedules by season, team, or both
sm_rounds Round data by season or ID
sm_markets List available betting markets
sm_bookmakers List available bookmakers

Using Includes (SportMonks)

SportMonks tools support an include parameter for enriched responses:

include: "participants,scores,league,statistics,events"
include: "player,position,teams"
include: "fixtures,rounds"

Using Filters (SportMonks)

Some endpoints support filters:

filters: "fixtureLeagues:8"       # Filter by league ID
filters: "eventTypes:14,18"        # Filter event types

Example Prompts

Once configured, you can ask Claude things like:

  • "Get me tonight's NBA odds from DraftKings and FanDuel — spreads, moneylines, and totals"
  • "Pull EPL standings for this season"
  • "What are the live soccer scores right now?"
  • "Compare the historical odds movement for the Lakers game last Tuesday"
  • "Get Arsenal's xG data and current squad"
  • "Show me value bets from SportMonks predictions for today's fixtures"
  • "Get player props available for tonight's NFL game"

Architecture

src/
├── index.ts          # MCP server entry — stdio transport (Claude Desktop / Claude Code)
├── server-http.ts    # MCP server entry — HTTP/SSE transport (claude.ai web)
├── odds-api.ts       # The Odds API tool definitions (10 tools)
├── sportmonks.ts     # SportMonks tool definitions (18 tools)
└── http.ts           # Shared HTTP utility (fetch + URL builder)

License

MIT

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选