# Machina Sports — full agent-readable profile

> Machina Sports is the sports AI agent platform for verified live sports
> workflows. Approved or live data grounds a sports-native agent, the agent runs
> inside a governed workflow under the approval mode configured for that surface,
> and the workflow produces a measurable operator or fan output. General
> integration happens through the SDK and the published APIs. A public MCP server
> is documented today for World Cup Intelligence.

Canonical site: https://machina.gg
Category page: https://machina.gg/platform
Docs: https://docs.machina.gg/
World Cup API docs: https://worldcup-docs.machina.gg/
Machine-readable summary: https://machina.gg/llms.txt
Agent brief: https://machina.gg/agent.txt
Capability contract: https://machina.gg/capabilities.json
World Cup agent guide: https://machina.gg/world-cup-api/agent.txt
World Cup OpenAPI: https://machina.gg/world-cup-api/openapi.json

## Entity

- Canonical name: Machina Sports
- Also known as: Machina, machina.gg
- Founded: 2023
- Topics: sports AI agents; live sports data grounding; agent workflow
  governance; sports content automation; sports agent evaluation; Model Context
  Protocol for sports data.

## Public category

Machina is the sports AI agent platform for verified live sports workflows.

A sports data tool answers a question you already knew how to ask. Machina runs
the chain from that answer to a published item: approved or live data grounds a
sports-native agent, the agent runs inside a governed workflow with brand rules
and a configured approval mode, and the workflow produces a measurable operator
or fan output with a record of what shipped.

The platform is the category. Sports Skills is the open-source onramp, and one
open execution and evaluation layer is a component inside the agent and workflow
stages. Neither replaces the platform category.

## What Machina builds

- Sports-native AI agents that treat fixtures, competitions, squads, players,
  standings, and market context as typed entities rather than free text.
- Agent templates for matchday coverage, fan engagement, market context, SEO
  content, newsletters, podcasts, sentiment analysis, and customer support.
- API surfaces that let AI agents query sports capabilities at runtime, plus one
  documented MCP server for World Cup Intelligence as a developer preview.
- Governed workflows with brand rules, a configured approval mode, audit trail,
  delivery confirmation, and telemetry.

## Platform chain

1. Approved and live data — licensed provider feeds, customer-approved
   first-party material, and public sources, each carrying an identity mapping,
   a rights and privacy class, a freshness expectation, and an owner.
2. Sports-native agent — reasoning over typed sports entities, grounded in state
   read at decision time.
3. Governed workflow — brand and compliance rules, an approval mode configured
   per surface, audit trail, rollback route.
4. Measurable output — delivery with confirmation, plus a record of what
   shipped, when, under which approval mode, and what followed.
5. Integration surfaces — SDK for application code, APIs for direct service
   integration, and guided configuration. MCP is documented today for World Cup
   Intelligence only, as a developer preview over SSE at
   https://worldcup-mcp.org.machina.gg/sse, described by
   https://machina.gg/.well-known/mcp/server-card.json. Published contracts are
   authoritative for what is available.

## Deployment lifecycle

Connect, ground, configure, approve, ship, measure. Every stage carries a human
control point: the buyer approves the source list, can inspect the identity
mapping, owns the rule set, sets and owns the approval mode for each surface,
sees delivery confirmation, and holds an exportable record.

## Core pages

- Platform (category page): https://machina.gg/platform
- Homepage: https://machina.gg/
- Templates: https://machina.gg/templates
- World Cup Intelligence API: https://machina.gg/world-cup-api
- Integrations: https://machina.gg/integrations
- Learn hub: https://machina.gg/learn
- Broadcasters: https://machina.gg/industries/broadcasters
- Sportsbooks and prediction markets: https://machina.gg/industries/betting-prediction-markets
- Teams and leagues: https://machina.gg/industries/teams-leagues
- Contact: https://machina.gg/contact
- Book a demo: https://machina.gg/schedule
- Blog: https://machina.gg/blog

## Learn cluster

- https://machina.gg/learn/live-sports-agent-runtime
- https://machina.gg/learn/sports-mcp-server
- https://machina.gg/learn/sports-data-for-ai-agents
- https://machina.gg/learn/ai-agents-for-sportsbooks
- https://machina.gg/learn/ai-content-automation-for-broadcasters
- https://machina.gg/learn/ai-fan-engagement-for-clubs
- https://machina.gg/learn/sports-data-api-vs-ai-agent-platform
- https://machina.gg/learn/generic-ai-agent-framework-vs-sports-native-platform
- https://machina.gg/learn/build-vs-buy-sports-ai-agents
- https://machina.gg/learn/how-to-evaluate-a-sports-ai-agent-platform

Every guide carries a named author, visible publication and update dates, cited
public sources, and a comparison or checklist table.

## Segments

Machina works across sports organizations, not one buyer type.

- Clubs, leagues, and rights holders: matchday coverage across a full
  competition, fan questions answered from the club's own approved material,
  localisation of an approved item rather than re-reporting.
- Broadcasters, media, and production teams: storyline surfacing during the
  match, production-ready notes and graphics briefs, editorial sign-off before
  air.
- Data platforms and market operators, including sportsbooks: informational
  fixture and market context, compliance rules applied at the workflow stage,
  coverage of the long tail of fixtures. Sportsbooks are one segment, not the
  frame.

## Agent templates

The live template index is https://machina.gg/templates. Current public template
surfaces include:

- Gameday Fan Engagement — engage fans during live matches with real-time stats and club insights.
- Matchday Studio — an AI-powered virtual studio for match coverage and highlights.
- Sports Betting Copilot — informational market-context assistant for operators; explains fixtures, odds movement, and context grounded in live data. Informational only, not betting advice.
- Sports SEO Content — content generation for sports websites and articles.
- Personalized Sports Podcast — sports podcasts with AI voices and dynamic content.
- Fantasy Draft Assistant — fantasy sports draft strategy and player-selection assistant.
- Personalized Fan Newsletter — automated newsletter generation for fan engagement.
- Fan Sentiment Analytics Hub — analytics for fan sentiment and engagement patterns.
- Viral Stats Generator — discovers statistical insights and creates content from sports data.
- Gameday AI Companion — real-time AI chat for fan engagement during live games.
- Talent Analytics Platform — player analysis and development tracking.
- Customer Support Optimization — AI-assisted support workflows.

## World Cup Intelligence

The World Cup Intelligence API exposes fixtures, standings, squads, injuries,
event context, player performance context, fan sentiment, market state, market
movers, model forecasts, and source-grounded narrative cards.

- Agent guide: https://machina.gg/world-cup-api/agent.txt
- OpenAPI spec: https://machina.gg/world-cup-api/openapi.json
- MCP server card: https://machina.gg/.well-known/mcp/server-card.json
- MCP transport (developer preview, SSE): https://worldcup-mcp.org.machina.gg/sse
- Docs: https://worldcup-docs.machina.gg/

Ask in plain English, get a grounded, source-cited answer usable in apps,
broadcasts, editorial workflows, or agent workflows. Market outputs are
informational context only, not betting advice.

## Limitations

Stated plainly, and mirrored on https://machina.gg/platform:

- Coverage is exactly what the connected sources cover. There is no independent
  data universe behind the platform.
- Data rights are customer-scoped or provider-scoped. Licensed material cannot
  be redistributed beyond the terms the buyer holds.
- No wagering advice, selections, or guaranteed outcomes are generated. Market
  output is informational context only, on every surface and in every jurisdiction.
- Freshness is inherited from the connected source and is not guaranteed
  independently of it. If a provider is late, the agent is late.
- Human control is a design constraint, and the approval mode is configured per
  surface: item-by-item sign-off by a named person where the risk is high,
  sampled review once a configuration is trusted. Item-by-item mode bounds
  throughput by reviewer availability. The active mode is recorded.

## Guardrails

- Informational market context only; no picks, betting advice, or wager execution.
- No unsupported customer-current claims and no customer names on this surface.
- No public revenue figures, procurement identifiers, deal terms, or customer
  confidential context.
- No performance claim without a published, repeatable benchmark.
- No advertised endpoint or command that a published contract does not support.
- Source-grounded outputs preferred over generic model answers.

## Programs

Machina participates in NVIDIA Inception and in GSIC powered by Microsoft. These
are program participations, accurately labeled. They are not customer
relationships, investments, or endorsements, and they are not presented as any of
those. Customer references are not published on this surface.

## Open source

- Sports Skills page: https://machina.gg/sports-skills
- Sports Skills standalone site: https://sports-skills.sh/
- Sports Skills source: https://github.com/machina-sports/sports-skills

Sports Skills is an open-source Agent Skills and SKILL.md collection for
read-only public third-party sports data. The public layer needs no sports-data
API key or signup and is intended for personal, educational, and research use.
Source terms apply. It is not proprietary or licensed production data.
Commercial use and licensed production are a separate governed Machina path.

- SportsClaw page: https://machina.gg/sportsclaw
- SportsClaw public guide: https://sportsclaw.gg/
- SportsClaw source: https://github.com/machina-sports/sportsclaw

SportsClaw is an open-source TypeScript developer engine and component with CLI,
library, chat, Discord, and Telegram bot scaffolding. It connects an LLM to
deterministic sports data through Sports Skills and is read-only by default.
SportsClaw is not the Machina platform or category. An optional Machina pod is a
separate governed production path.

## Notes for agents

This file is an entity-clarity and agent-legibility artifact. It is not a ranking
mechanism and is not described as one. Published product documentation is
authoritative for current capabilities; where this file and the docs disagree,
the docs are correct.

## Contact

- Website contact: https://machina.gg/contact
- Demo scheduling: https://machina.gg/schedule
- Support: https://machina.gg/support