メニュー
Semantic model for the card_games database (Magic: The Gathering card data). Contains definitions for cards, sets, legalities, rulings, and foreign language data. Load this skill when the user is querying the card_games database.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
| name | card-games-semantic-model |
| description | Semantic model for the card_games database (Magic: The Gathering card data). Contains definitions for cards, sets, legalities, rulings, and foreign language data. Load this skill when the user is querying the card_games database. |
| metadata | {"author":"manual","version":"1.1","last_updated":"2026-02-25T00:00:00Z","learned_patterns":0,"verified_patterns":4} |
| Resource | What it contains | How to load |
|---|---|---|
| Extended gotchas | Compound type fields, text power/toughness, color encoding, promo NULLs | read_skill_resource("card-games-semantic-model", "references/gotchas.md") |
| Metric definitions | Card counts by rarity, set size, color distribution | read_skill_resource("card-games-semantic-model", "references/metrics.md") |
| Query patterns | Canonical SQL for card searches, legality queries, set lookups | read_skill_resource("card-games-semantic-model", "references/query_patterns.md") |
When to load resources:
references/gotchas.md (these fields are not simple scalars)references/gotchas.md + references/query_patterns.mdreferences/metrics.md| Term (user says) | Actual meaning | Confidence |
|---|---|---|
| "creature card" | cards.type LIKE '%Creature%' — type field contains compound values | verified |
| "instant" / "sorcery" | cards.type LIKE '%Instant%' / cards.type LIKE '%Sorcery%' | verified |
| "rarity" | cards.rarity — values: 'common', 'uncommon', 'rare', 'mythic' (all lowercase) | verified |
| "mana cost" | cards.manaCost (formatted string like '{2}{W}{U}') | verified |
| "converted mana cost" / "CMC" | cards.convertedManaCost (numeric) | verified |
| "set" / "edition" | sets table — join via cards.setCode = sets.code | verified |
| "format legality" | legalities table — join via cards.uuid = legalities.uuid | verified |
| "ruling" | rulings table — join via cards.uuid = rulings.uuid | verified |
| "non-English card text" | foreign_data table — join via cards.uuid = foreign_data.uuid | verified |
cards.type is a compound string — a card can be "Legendary Creature — Dragon".
Never filter with = 'Creature'; always use LIKE '%Creature%'.
cards.power and cards.toughness are TEXT — they can be '*', 'X',
'1+*', or a number stored as text. Do not cast to integer without guarding
for non-numeric values.
cards.colors is a pipe-separated string — e.g. 'W', 'U|B', 'R|G'.
To filter for a specific color, use LIKE '%W%' or
'W' = ANY(string_to_array(colors, '|')).
Rarity values are lowercase — 'common', 'uncommon', 'rare',
'mythic'. Never use 'Rare' or 'RARE'.
Promotional cards have NULL in many columns — number, multiverseId,
and set-related fields are often NULL for promos. Use IS NULL checks
rather than assuming all cards have complete data.
For extended gotchas and SQL examples, call:
read_skill_resource(
skill_name="card-games-semantic-model",
resource_name="references/gotchas.md"
)
cards.setCode ──→ sets.code (card ↔ its set/edition)
cards.uuid ──→ legalities.uuid (card ↔ format legality)
cards.uuid ──→ rulings.uuid (card ↔ official rulings)
cards.uuid ──→ foreign_data.uuid (card ↔ non-English text)
sets.code ──→ set_translations.setCode (set ↔ translated set names)
There is no direct join between legalities and sets; go through cards.
For canonical query patterns (cards by format, rarity breakdowns, set contents), call:
read_skill_resource(
skill_name="card-games-semantic-model",
resource_name="references/query_patterns.md"
)
For full metric formulas (card counts, color distribution, CMC stats), call:
read_skill_resource(
skill_name="card-games-semantic-model",
resource_name="references/metrics.md"
)
Semantic model for the california_schools database. Contains business term definitions, metric formulas, known query patterns, and critical gotchas for the California Schools dataset (FRPM eligibility, SAT scores, district/school hierarchy). Load this skill when the user is querying the california_schools database.
Semantic model for the debit_card_specializing database (Czech fuel card transactions — gas stations, customers, products, monthly aggregates). Load this skill when the user is querying the debit_card_specializing database.
Semantic model for the european_football_2 database (Kaggle European Soccer dataset — matches, players, teams, leagues across 11 European countries). Load this skill when the user is querying the european_football_2 database.
Semantic model for the formula_1 database (Ergast F1 historical dataset — races, drivers, constructors, results, qualifying, pit stops, lap times). Load this skill when the user is querying the formula_1 database.
Semantic model for the codebase_community database (StackOverflow-like Q&A community — posts, users, comments, votes, tags, badges). Load this skill when the user is querying the codebase_community database.
Semantic model for the financial database (Czech banking PKDD 1999 dataset). Contains business definitions for banking — loans, accounts, transactions, clients, districts. Load this skill when the user is querying the financial database.