菜单
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.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
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 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 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.
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.
基于 SOC 职业分类
| name | european-football-2-semantic-model |
| description | 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. |
| 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 | API IDs vs primary keys, time-series attribute joins, home/away goal logic | read_skill_resource("european-football-2-semantic-model", "references/gotchas.md") |
| Metric definitions | Win rates, average goals per match, player ratings, form metrics | read_skill_resource("european-football-2-semantic-model", "references/metrics.md") |
| Query patterns | Team standings, top-rated players, head-to-head records | read_skill_resource("european-football-2-semantic-model", "references/query_patterns.md") |
When to load resources:
references/gotchas.md (API ID vs id is critical)Player_Attributes or Team_Attributes → load references/gotchas.md (time-series trap)references/gotchas.md + references/metrics.md| Term (user says) | Actual meaning | Confidence |
|---|---|---|
| "home team" | Match.home_team_api_id (join to Team.team_api_id) | verified |
| "away team" | Match.away_team_api_id (join to Team.team_api_id) | verified |
| "overall rating" | Player_Attributes.overall_rating (0–100 scale) | verified |
| "season" | Match.season — stored as text: '2008/2009', '2009/2010', etc. | verified |
| "league" | League.name — join: Match.league_id → League.id | verified |
| "country" | Country.name — join: League.country_id → Country.id | verified |
| "team rating" | Team_Attributes.buildUpPlaySpeed / overall etc. — time-series, filter by date | verified |
| "lineup" | Match.home_player_1 … Match.home_player_11 — store player_api_id values | verified |
Always join on team_api_id / player_api_id, NOT id — the id
column is an internal SQLite/PostgreSQL row ID. The stable cross-table key
is team_api_id for teams and player_api_id for players.
Player_Attributes and Team_Attributes are time-series — each player/team
has multiple rows across different dates. Joining without a date filter creates
a Cartesian product. Use a subquery to get the most recent record.
A team's total goals requires combining home and away — Match stores
home_team_goal and away_team_goal separately. You need a CASE expression
to aggregate from both perspectives.
Match.season is a text field — format is '2008/2009'. Filter with
= '2008/2009', not = 2008 or = 2009.
Match has 11 player columns per side — home_player_1 through
home_player_11 each store a player_api_id. There is no normalised
player-per-match table; lineup queries require unpivoting these columns.
For extended gotchas and SQL examples, call:
read_skill_resource(
skill_name="european-football-2-semantic-model",
resource_name="references/gotchas.md"
)
Match.home_team_api_id ──→ Team.team_api_id (match ↔ home team)
Match.away_team_api_id ──→ Team.team_api_id (match ↔ away team)
Match.league_id ──→ League.id (match ↔ league)
League.country_id ──→ Country.id (league ↔ country)
Player_Attributes.player_api_id ──→ Player.player_api_id (attributes ↔ player)
Team_Attributes.team_api_id ──→ Team.team_api_id (attributes ↔ team)
Match.home_player_1..11 ──→ Player.player_api_id (lineup ↔ player)
There is no direct join between Match and Player_Attributes. Go through
Player via player_api_id.
For canonical query patterns (season standings, top scorers, player ratings), call:
read_skill_resource(
skill_name="european-football-2-semantic-model",
resource_name="references/query_patterns.md"
)
For full metric formulas (goals per game, win rates, player rating trends), call:
read_skill_resource(
skill_name="european-football-2-semantic-model",
resource_name="references/metrics.md"
)