TCG-SCHEMA is an open, schema.org-aligned ontology for Trading Card Games (TCG/CCG/LCG). It provides a structured semantic layer for:

• Cards vs printings
• Sets and releases
• Decklists and formats
• Mechanics and counters
• Resource systems (mana, ink, energy, etc.)
• Metagame archetypes and matchups
• Evidence-based counterclaims
• LLM-friendly card descriptors

The goal is simple and ambitious:

Make card games machine-readable without reducing them to a toy simulator.

Most card game data exists in one of three forms:

1. Raw card databases (names, text, sets)
2. Market data (prices, printings, SKUs)
3. Decklists and tournament results

What’s missing is a formal semantic layer that connects:

• Game rules
• Card intent and interaction axes
• Deckbuilding roles
• Format legality
• Metagame structure
• AI-assisted interpretation

tcg-schema fills that gap.

A tcg:Card is the abstract identity. A tcg:CardPrinting is a specific product instance.

This distinction is non-negotiable.

The core schema supports:

• Magic
• Lorcana
• Pokémon
• Flesh and Blood
• Yu-Gi-Oh
• Future games not yet invented

Game-specific modules extend the core (e.g., tcg-mtg).

We model structure and meaning — not full gameplay simulation.

This allows:

• Knowledge graphs
• Search
• Deckbuilding heuristics
• Counter-analysis
• LLM grounding

Without building a digital judge program.

Counter relationships are not dogma.

A tcg:CounterClaim must have:

• format
• time window
• sample size
• method
• evidence

Metagames drift. The graph must reflect that.

Cards include structured descriptors:

• role (removal, engine, finisher, etc.)
• interaction axis (stack, graveyard, combat, resource denial)
• synergy tags
• polarity (helpsOwnPlan / disruptsOpponent)
• confidence and extraction method

This enables explainable AI-assisted deck construction.

tcg-schema/
│
├── core/
│   └── tcg-schema-core.ttl
│
├── descriptors/
│   └── tcg-descriptors.ttl
│
├── mtg/
│   ├── tcg-mtg-extension.ttl
│   ├── mtg-rules-sections.ttl
│   └── mtg-terms.ttl
│
├── examples/
│   ├── card-example.jsonld
│   ├── deck-example.jsonld
│   └── counterclaim-example.jsonld
│
└── docs/
    └── site/

• tcg:CardGame
• tcg:CardSet
• tcg:Format
• tcg:Legality

• tcg:Card
• tcg:CardPrinting
• tcg:CardFace
• tcg:Ability
• tcg:Effect
• tcg:CounterType

• tcg:ResourceSystem
• tcg:ResourceType
• tcg:ResourceCost
• tcg:ResourceCostComponent

Works for:

• Mana
• Ink
• Energy
• Pitch systems
• Hybrid systems

• tcg:Archetype
• tcg:Matchup
• tcg:CounterClaim
• tcg:EvidenceBundle

• tcg:DescriptorType
• tcg:CardDescriptor
• tcg:ExtractionMethod
• tcg:DescriptorEvidence

• A format-aware deck builder from your collection
• A metagame analysis engine
• A cost-effective counter recommender
• A semantic search engine for card interactions
• A structured RAG pipeline grounded in card rules
• A knowledge graph across multiple TCGs

A card:

• has role = removal
• interactionAxis = stackInteraction
• polarity = disruptsOpponent
• costProfile = cheap
• confidence = 0.92

A counterclaim:

• archetype A counters archetype B
• winRate = 0.58
• sampleSize = 432
• timeWindow = Jan–Mar 2026
• method = logisticRegression

No vibes. Just structured evidence.

Early core is stable. Game-specific modules are in progress.

Planned:

• MTG complete rule anchoring
• Lorcana extension
• Pokémon extension
• JSON-LD examples
• SHACL validation shapes
• Public SPARQL endpoint (experimental)

We welcome:

• Ontology feedback
• New game modules
• Descriptor vocabulary proposals
• Validation shapes (SHACL)
• Dataset alignment mappings
• Example JSON-LD documents
• Documentation improvements

Open an issue. Be precise. Cite rules when relevant.