# Explore tables and columns before you write SQL

Source: https://www.hotdata.dev/use-cases/table-catalog
Site index: https://www.hotdata.dev/llms.txt

Different databases usually mean different tools. Here you list connections, inspect tables and columns, and refresh metadata after upstream changes. The workflow stays the same whether you're on Postgres, Snowflake, or something else.

## How it works

### Step 1: Pick the connection

**Claude**

```
What data sources are connected in this workspace, and what short name should I use for each when I run commands?
```

**CLI**

```bash
hotdata ingest sources list
```

**Python SDK**

```python
import hotdata

connections = hotdata.ConnectionsApi(api_client)
connections.list_connections()
```

### Step 2: Inspect tables and columns

**Claude**

```
On our finance database, list tables whose names start with “revenue.” For each one I need every column: its name, type, and whether blank values are allowed.
```

**CLI**

```bash
hotdata databases tables list --table revenue% --schema public
```

**Python SDK**

```python
import hotdata

api = hotdata.InformationSchemaApi(api_client)
api.information_schema(
    connection_id="finance",
    table="revenue%",
    include_columns=True,
)
```

### Step 3: After the database schema changes

**Claude**

```
Finance changed their database structure. Refresh my catalog here so what I see matches what’s actually there now.
```

**Python SDK**

```python
import hotdata

refresh = hotdata.RefreshApi(api_client)
refresh.refresh(hotdata.RefreshRequest(connection_id="finance"))
```

A schema refresh reconciles the catalog in the background. Let it settle before **`hotdata databases tables list`** reflects brand-new DDL.

### Step 4: Page through a wide catalog

**Claude**

```
Give me a browseable snapshot: up to 50 tables from finance with full column details so I can skim what exists.
```

**CLI**

```bash
hotdata databases tables list --limit 50
```

**Python SDK**

```python
import hotdata

api = hotdata.InformationSchemaApi(api_client)
api.information_schema(
    connection_id="finance",
    include_columns=True,
    limit=50,
)
```

## Who uses this

- Anyone writing SQL or filters who wants names and types confirmed first.
- Engineers building table pickers on top of workspace catalog output.
- Teams checking renames before dashboards ship.
- Agent tooling that should only touch columns the catalog actually lists.