agentscope-ai/schema-exploration

Schema Exploration Skill

When to Use This Skill

Use schema-exploration when the user:

• Asks "what tables are in the database?"
• Asks "what columns does the X table have?"
• Asks about relationships between tables
• Needs to understand data types before writing a query
• Wants an entity-relationship overview

Workflow

Step 1 — List All Tables

Use sql_list_tables to see everything available.

Tool: sql_list_tables
(no parameters)

Step 2 — Inspect Relevant Tables

Use sql_get_schema with the table name(s) you need to understand:

Tool: sql_get_schema
tables: "Customer"           # single table
tables: "Invoice,Customer"   # multiple tables at once

This returns:

• Columns — names, types, NOT NULL constraints, primary keys
• Foreign keys — links to other tables
• Sample data — 3 example rows to understand content

Step 3 — Map Relationships

Look for columns ending in Id — they are almost always foreign keys.

Example for the Chinook database:

Customer.SupportRepId → Employee.EmployeeId
Invoice.CustomerId    → Customer.CustomerId
Track.AlbumId         → Album.AlbumId
Track.GenreId         → Genre.GenreId

Step 4 — Present the Findings

Provide:

• A list of all tables with a one-line purpose summary
• Column names and types for any tables the user asked about
• The relationship chain (which table links to which)
• Sample data to illustrate what the table holds

---

Examples

"What tables are available?"

1. Call sql_list_tables
2. Return a formatted list with brief descriptions:

The Chinook database has 11 tables:

Music Catalog:
  - Artist       — 275 music artists
  - Album        — 347 albums (linked to Artist)
  - Track        — 3,503 tracks (linked to Album, Genre, MediaType)
  - Genre        — 25 music genres
  - MediaType    — 5 file formats (MP3, AAC, …)

Commerce:
  - Customer     — 59 customers from 24 countries
  - Invoice      — 412 purchase invoices
  - InvoiceLine  — 2,240 line items (linked to Invoice and Track)

Staff:
  - Employee     — 8 employees with reporting hierarchy

Playlists:
  - Playlist      — 18 playlists
  - PlaylistTrack — join table linking Playlist ↔ Track

"What does the Customer table look like?"

1. Call sql_get_schema with tables: "Customer"
2. Present columns with types and notes:

Customer table:
  CustomerId   INTEGER  PK
  FirstName    TEXT
  LastName     TEXT
  Company      TEXT     (nullable)
  Address      TEXT
  City         TEXT
  State        TEXT     (nullable)
  Country      TEXT
  PostalCode   TEXT     (nullable)
  Phone        TEXT     (nullable)
  Fax          TEXT     (nullable)
  Email        TEXT
  SupportRepId INTEGER  FK → Employee.EmployeeId

Sample rows show customers from Brazil, Germany, and Canada.

"How are artists connected to sales?"

Map the full JOIN path:

Artist (ArtistId)
  ↓ 1:many
Album (ArtistId, AlbumId)
  ↓ 1:many
Track (AlbumId, TrackId)
  ↓ 1:many
InvoiceLine (TrackId, UnitPrice, Quantity)
  ↓ many:1
Invoice (InvoiceId, Total)

Then suggest using the query-writing skill to write the aggregation query.