Vublox MCP Server – Live Football for AI Agents

TL;DR

TL;DR: Vublox MCP Server lets AI agents query live football scores, player profiles, and match summaries through the Model Context Protocol — no API key needed.

Source and Accuracy Notes

⚠️ This section is MANDATORY. All links must be verified from actual source, not guessed.

• Project page: vublox.com ← live product site
• Source repository: github.com/Vublox/vublox-agent-tools ← README read in full
• License: MIT (verified via package.json "license": "MIT")
• HN launch thread: not available at time of writing
• Source last checked: 2026-07-11 (commit main branch)

What Is Vublox MCP Server?

The Vublox MCP Server is a Model Context Protocol server that bridges AI agents to live football data on Vublox. Once installed in an MCP-compatible client (Claude Desktop, VS Code Copilot, etc.), you can ask your agent natural-language questions about live matches, scores, and player information — and it will fetch structured data from Vublox’s backend.

The key differentiator is that it requires no API key. The server handles the connection to Vublox directly, so you just install it and query.

Available Tools

The server exposes six MCP tools:

| Tool | What it returns | |------|-----------------| | search_sports_events | Matches by team, league, or keyword | | search_players | Football players by name with profile links | | get_live_matches | All currently live football matches | | get_match_summary | Full match info for a specific match | | get_todays_goals | Recent scores across leagues (last 48 hours) | | get_recent_matches | Finished matches from the last 48 hours |

Setup Workflow

Step 1: Install for Claude Desktop

Add the following to your Claude Desktop config file (typically ~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "vublox-sports": {
      "command": "npx",
      "args": ["@vublox/agent-tools"]
    }
  }
}

Then restart Claude Desktop.

Step 2: Install for VS Code Copilot

Add to your .vscode/mcp.json (create it if it does not exist):

{
  "servers": {
    "vublox-sports": {
      "type": "stdio",
      "command": "npx",
      "args": ["@vublox/agent-tools"]
    }
  }
}

Step 3: Query Your Agent

After the server is connected, try these prompts in Claude or Copilot:

• “What football matches are live right now?”
• “How did Arsenal do today?”
• “Show me today’s Premier League scores”
• “Who is Kylian Mbappé?”
• “Tell me about the Barcelona vs Real Madrid match”

Deeper Analysis

Use case: This MCP server is aimed at developers building AI assistants that need sports context — for example, a news bot that references today’s scores, a fantasy football advisor, or a sports community chatbot.

Technical footprint: The server is a TypeScript project using the official @modelcontextprotocol/sdk. It targets Node 18 or higher and ships compiled JavaScript via a dist/ folder. No external API key is needed because it connects directly to Vublox.

Package: @vublox/agent-tools at version 0.1.5. Published to npm.

Strengths:

• Zero-config setup — no API key, no registration
• Covers live matches, recent results, and player lookups
• Works with any MCP-compatible client

Limitations:

• No license file in the repository (MIT declared in package.json only)
• No GitHub stars or community traction yet (0 stars, 0 forks at time of writing)
• Relies on Vublox being available — no offline fallback

Security Notes

• No API key is transmitted — queries go directly to Vublox
• The server communicates over stdio with the host agent, keeping data local to that process
• As with any MCP server, only install from sources you trust

FAQ

Q: Does this work with ChatGPT? A: Not directly. ChatGPT does not currently support the MCP stdio protocol. This server is designed for MCP-compatible clients such as Claude Desktop, VS Code Copilot, and custom MCP integrations.

Q: Is an API key required? A: No. The server connects to Vublox directly without authentication.

Q: What leagues and competitions are covered? A: The Vublox platform covers major football leagues. The exact coverage is determined by Vublox — check vublox.com for the full list of supported competitions.

Q: Can I self-host this MCP server? A: Yes. Clone the repo, run npm install && npm run build, then point your MCP client to node dist/index.js.

Conclusion

The Vublox MCP Server fills a specific niche — giving AI agents live sports data through a familiar protocol. For developers building sports-aware assistants or community bots, it removes the friction of finding and paying for a sports API. Setup takes under five minutes in Claude Desktop or VS Code Copilot, and no credentials are needed.

If you want a clean way to bring real-time football context into an AI workflow, this is one of the simplest options available right now.