Schema Explorer

# Schema Explorer

Progressively reveal structure through layered discovery.

## Category
Tool Discovery

## Context
Working with complex data structures.

## Problem
Exposing full schema upfront is overwhelming; too little is insufficient.

## Solution
Layer discovery progressively:
- **Layer 1**: list_tables() - What exists
- **Layer 2**: describe_table(name) - Field details
- **Layer 3**: sample_data(table, limit) - See examples
- **Layer 4**: get_query_hints(table) - Usage tips

## Example (Python)
```python
# Layer 1: What tables exist?
@tool
def list_tables() -> list[str]:
    """List all database tables."""

# Layer 2: What fields does a table have?
@tool
def describe_table(table_name: str) -> TableSchema:
    """Get schema for a table. Call list_tables() first."""

# Layer 3: What does the data look like?
@tool
def sample_rows(table_name: str, limit: int = 5) -> list[dict]:
    """Get sample rows. Call describe_table() first."""
```

## Considerations
- Each layer should suggest the next step
- Include examples at each level
- Consider caching schema information

## Related Patterns
- discovery-tool
- dependency-hint
- progressive-detail

---
Source: https://arcade.dev/patterns/schema-explorer