Sports intelligence
for AI agents

SportMind gives AI agents the intelligence, reasoning, and context to understand sports — the commercial, financial, and competitive signals the industry runs on.

744
Skill files
130
Calibration records
96%
Direction accuracy
21
Sports validated

The problem

AI agents don't know sports.

A general-purpose LLM given a match to analyse produces generic output. It does not know the things that actually change outcomes. SportMind teaches agents exactly those things.

Cricket
An evening T20 in Mumbai with 78% humidity means the team batting second has a structural advantage. Dew accumulates. Spin bowling fails. Standard models miss this.
MMA
A fighter missing weight signals preparation failure, altered psychology, and physiological compromise. It is categorically different from a team losing a regular game.
Formula 1
A 0.3 second qualifying advantage predicts race outcomes more reliably than season form. But only on certain circuit types. SportMind knows which ones.
Fan tokens
With Fan Token™ Play, a WIN prediction is simultaneously a SUPPLY REDUCTION event. Tokens burn on wins — permanently. A loss mints tokens to treasury — supply increases. A standard model sees half the signal. SportMind models both.
World Cup
Not all tokens claiming national team affiliation are official. $ARG and $POR are confirmed Fan Tokens™. Hundreds of fraudulent tokens will proliferate around the 2026 World Cup. SportMind's verification framework teaches agents to tell the difference.

How it works

The engine needs a transmission.

AI agents are powerful. Without structured sports intelligence they produce generic output. SportMind is the transmission layer — converting raw sporting events into calibrated signals that agents can reason about and act on.

Data sources
Official club X/Twitter
Chiliz Chain on-chain
LunarCrush · X API
Press conferences
Macro / crypto
fantokens.com
chiliscan.com
Regulatory sources
SportMind
intelligence
Layer 5 · Macro
Layer 4 · Market
Layer 1 · Sport domain
Layer 2 · Athlete
Layer 3 · Fan token
Core intelligence
Reasoning architecture
Agent
patterns
Pre-match chain
FTP monitor
Post-match analysis
Governance delegate
Scouting agent
Output
Structured signal · JSON
Plain-English brief · fans
Squad brief · team news
Governance brief · votes
Scout report · transfers
Verification signal · fraud flags
Human
in the loop
Fan token position decision
Governance vote
Transfer committee
Application execution

SportMind agents produce intelligence. They do not execute trades, submit votes, or negotiate contracts. The human decision point is architectural — not a limitation.

The library

Six layers. Expanding intelligence.

Load in order — macro → market → sport domain → athlete → fan token → deployment. The six-layer loading architecture remains constant. The intelligence, reasoning, and context within grows continuously — building toward a complete Mind for Sports. Compatible with any LLM. Skills are structured markdown, not API wrappers. Each layer adds intelligence about the domain, reasoning frameworks for interpreting signals, and context for understanding why outcomes happen — not just what happened.

Layer 1
Sport domain
42 sports. Event playbooks, risk variables, signal weights, and agent reasoning prompts. Six Tier A sports with full fan token intelligence integration.
sports/
Layer 2
Athlete intelligence
45 files across active sports. Form, availability, composite modifier (0.55–1.25×).
athlete/
Layer 3
Fan token commercial
21 skills. Lifecycle phases 1–6, Fan Token Play (PATH_2 confirmed), DeFi, governance, RWA, tournament exit intelligence.
fan-token/
Layer 4
Market intelligence
48 documents. Commercial tier, fanbase depth, competition calendar, sports equity signals, and exchange intelligence.
market/
Layer 5
Macro intelligence
20 documents. Regulatory frameworks across UK, US, EU, APAC, and MENA. Crypto cycles. Institutional intelligence. Exchange signals. Government strategy. Stablecoin and CBDC.
macro/
Layer 6
Deployment intelligence
6 skills. Telegram delivery — sentiment monitoring, price explainers, pre-match signals, macro interpretation.
telegram/

Developer toolkit

The library is the intelligence.
The toolkit is how you deploy it.

Copy-paste templates, 22 agent prompts, and 11 agentic workflow patterns. Three concrete examples of what the toolkit produces:

Pattern 10 — Scouting Agent
Ranks transfer targets by commercial value-to-fee ratio. The formula that powers it:
CVS = APS×0.30 + AELS×0.25 + DTS×0.20 + PI×0.15 + LTUI×0.10 FAS = CVS / log₁₀(fee_m + 1)
Output: EXCELLENT / GOOD / MODERATE / POOR tier per target
Prompt 21 — Fan-Facing Brief
Plain English pre-match summary — no jargon, five sections, includes a Fan Token Play note for $AFC holders. Built for fans, not developers.
"Arsenal are slight favourites today. Saka is doubtful — that matters. The $AFC supply burn only happens on a win. If Arsenal lose, supply stays neutral."
Prompt 21 in agent-prompts/agent-prompts.md
Pattern 8 — FTP Monitor
Full Fan Token Play match cycle: T-48h pre-liquidation detection → kickoff → T+48h settlement. Prevents the most common agent error on PATH_2 tokens.
T-48h: pre_liquidation detected → PROTOCOL_EVENT (never bearish) WIN: burn confirmed on-chain ✓ LOSS: mint_to_treasury — supply increases
examples/agentic-workflows/fan-token-play-monitor.md

The library extends beyond these six core layers toward a complete Mind for Sports — fourteen dimensions of Intelligence, Reasoning, Context, Memory, Judgment, Attention, Learning, Integration, Communication, Calibration, Adaptation, Verification, Ethics, and Transparency. Additional intelligence domains include psychological and coaching intelligence, blockchain and on-chain reasoning, odds and market intelligence, sports integrity, stablecoin and CBDC frameworks, fraud verification, and more.

Full intelligence map →

Signal output

Every signal includes not just a direction — but the reasoning chain and context behind it.

What SportMind produces.

Every analysis returns a consistent structured output — direction, confidence score, modifiers applied, and active flags.

sportmind · pre-match signal ⚽ football · ucl
{
  "direction":           "HOME",
  "adjusted_score":      72.4,
  "sms":                 79,
  "recommended_action":  "ENTER",
  "composite_modifier":  1.10,
  "confidence_level":    "HIGH",
  "signal_class":        "EXECUTION",

  "modifiers_applied": {
    "athlete_modifier":     1.10,
    "macro_modifier":       1.00,
    "venue_modifier":        1.05,
    "officiating_modifier":  1.02
  },

  "flags": {
    "lineup_unconfirmed":    false,
    "ftp_path2_active":      true,
    "supply_event_type":     "REDUCTION",
    "macro_override_active": false
  }
}
PSG vs Arsenal · UCL · Puskás Arena, Budapest v3.97.113

Empirical foundation

Every claim is verifiable.

130 records across 21 sports. All in the repository. Submitted before real matches. Including the wrong ones.

96%
125 of 130 records correct
Direction accuracy 125 / 130
Zero wrong-direction records
qualifying_delta (F1)4/4 ✓
dew_factor (cricket)5/5 ✓
india_pakistan ×2.003/3 ✓
morning_skate (NHL)3/3 ✓
raider_primacy (kabaddi)1/1 ✓
21 sports calibrated
Tier A — fan token active
football cricket basketball mma formula1 motogp
Tier B & C — monitored
tennis ice-hockey rugby-union rugby-league afl athletics snooker darts swimming winter-sports kabaddi handball nascar netball rowing

Who uses SportMind

Built for every builder.

SportMind is the intelligence, reasoning, and context layer — the foundation any AI-powered sports application needs and none of them should have to build from scratch. The possibilities are endless.

Community

Calibration leaderboard.

The first 10 external contributors become Founding Calibrators — permanently credited in the library's history. Every record moves a modifier forward.

Rank Contributor Records Status
1 SportMind 130 · 96% Maintainer
2 Open Founding Calibrator
3 Open Founding Calibrator
4 Open Founding Calibrator
5 Open Founding Calibrator
6 Open Founding Calibrator
7 Open Founding Calibrator
8 Open Founding Calibrator
9 Open Founding Calibrator
10 Open Founding Calibrator
Read the First Record Challenge →

Verification

Know what is official.

Not all sports crypto assets are equal. SportMind maintains a verified registry of official Fan Tokens™ and a fraud risk intelligence framework to help agents and holders distinguish officially licensed tokens from unverified assets.

Method
Four-source verification methodology. Club official → socios.com → chiliscan.com → fantokens.com. All four required.
Red flags
Three-tier red flag framework. Tier 1: any single flag triggers UNVERIFIED. Passive yield promises are a critical dual signal — fraud indicator and CLARITY Act violation.
Tournaments
World Cup fraud amplification signals. Fraudulent national team tokens proliferate in the four weeks before and during major tournaments. The risk is highest during World Cup 2026.
Read the verification framework →

Contribute

Every record moves a modifier forward.

The first 10 external contributors become Founding Calibrators — permanently credited in the library's history.

Submit a calibration record
Run SportMind before a real match. Submit the outcome. No coding required. Wrong predictions are as valuable as correct ones.
30 min
★★
Expand a stub sport
Several sports have lighter coverage and benefit from domain expertise — badminton, volleyball, table tennis, and more.
2–3 hrs
★★★★
External recalibration
Submit 10+ records for one modifier and write the analysis. First external voice in SportMind's calibration history.
8+ hrs
Read the First Record Challenge →

Exploring a use-case not listed here? SportMind is an open playground. See all use-cases →

The SportMind Suite

Ready-to-deploy kits.

Opinionated starting points built on the SportMind intelligence layer. Fork, configure, and ship.

Fan Token Wallet Kit
Fan Token Agentic Wallet Starter Kit
Deploy an agentic wallet with SportMind intelligence built in. Fan token positions, on-chain signals, and match-driven triggers — ready to fork.
On-chain Chiliz Chain Forkable
Telegram Bot Kit
Telegram AI Bot Starter Kit
Launch a SportMind-powered Telegram bot in minutes. Pre-match signals, fan token sentiment, and macro intelligence — delivered where your community already is.
Managed Bots Forkable
Intelligence Agent
Intelligence Agent
Monitors public sources on a schedule, classifies signals against the SportMind intake framework, and delivers a structured briefing with reasoning. You decide what gets added to the library.
Scheduled RSS GitHub Issues
sportmind-automation
Coming soon
Automated library maintenance and signal pipeline. GitHub Actions plus Claude API.

Free to use.
Free to build on.

MIT licensed. Compatible with any LLM.
Intelligence, reasoning, and context — ready to deploy.