MCP for AI Agents.
The Model Context Protocol (MCP) lets AI assistants like Claude or ChatGPT call your sports data API as a tool — no manual API calls needed. Ask your AI "who plays tonight?" and it finds out itself.
What is MCP?
MCP (Model Context Protocol) is a standard that lets AI assistants call external tools in real time. When you connect our MCP server to Claude Desktop (or any compatible client), the AI can call tools like list_matches or get_prediction during a conversation — without you needing to copy-paste API responses. The AI reads the data and answers your question directly.
All of our sport MCP servers speak JSON-RPC 2.0 over HTTP POST. They use the same API keys as the REST API — no extra sign-up needed.
Authentication
Use the same Bearer token you use for the REST API. Pass it in the Authorization header:
Get a key at sports.bzzoiro.com/register/.
Connect to Claude Desktop
Add each sport you want to Claude Desktop's config file. The config lives at:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
Add one entry per sport under mcpServers. Here's a full example with all four sports:
After saving the file, restart Claude Desktop. You'll see each sport server listed in the tools panel.
Tennis MCP
| Tool | Description | Key inputs |
|---|---|---|
| list_matches | Get matches filtered by status, tournament or date | status, tournament_id, date_from, date_to |
| get_match | Full match detail with live score and set breakdown | match_id |
| list_players | Search players by name | q (search query) |
| get_player | Player profile with ranking and country | player_id |
| list_predictions | Win/set/game ML predictions for upcoming matches | upcoming (boolean) |
| get_prediction | Full prediction detail for one match | prediction_id |
| list_tournaments | Active ATP/WTA tournaments | active (boolean) |
| list_rankings | ATP or WTA world rankings | gender (m/w), limit |
CS2 MCP
| Tool | Description | Key inputs |
|---|---|---|
| list_matches | CS2 matches with map scores and BO format | status, tournament_id |
| get_match | Match detail with full map-by-map breakdown | match_id |
| list_teams | Search CS2 teams by name | q (search query) |
| get_team | Team profile with ELO, country and roster | team_id |
| list_players | Search CS2 players by name or team | q, team_id |
| get_player | Player KD ratio, damage, headshot rate | player_id |
| list_predictions | CatBoost win predictions | upcoming (boolean) |
| get_prediction | Full prediction for one match | prediction_id |
| list_tournaments | CS2 tournaments (ESL, BLAST, FACEIT, etc.) | active (boolean) |
Darts MCP
| Tool | Description | Key inputs |
|---|---|---|
| list_matches | PDC darts matches by status | status |
| get_match | Match detail with set and leg scores | match_id |
| list_players | All darts players with ELO and PDC ranking | limit, offset |
| get_player | Full player profile with recent form | player_id |
| list_predictions | ELO-based predictions with High/Medium/Low confidence | — |
| get_prediction | Prediction detail for one match | prediction_id |
| list_rankings | Current PDC world rankings | limit |
Hockey MCP
| Tool | Description | Key inputs |
|---|---|---|
| list_matches | Hockey matches with OT/SO flags and period scores | status |
| get_match | Full match detail including period breakdown | match_id |
| list_teams | All hockey teams | limit, offset |
| get_team | Team profile and recent results | team_id |
| list_predictions | ML win predictions for upcoming hockey matches | — |
| get_prediction | Prediction detail for one match | prediction_id |
HTTP transport
Each MCP endpoint accepts HTTP POST with a JSON-RPC 2.0 body. You don't need this section to use Claude Desktop — it's for developers building their own MCP clients.