Your task is to classify a natural language question into one of the following three categories, based on whether it can be answered using SQL over the given database schema.

Classification Categories:
1. **"answerable"** - The question can be clearly answered with the given database schema. All required information (tables, columns, relationships) exists in the schema.

2. **"ambiguous"** - The question is unclear, ambiguous, or requires clarification. This includes:
   - Questions with unclear references ("it", "that", "the previous one")
   - Questions with multiple possible interpretations without further clarification
   - Questions that are too vague to understand the intent

3. **"unanswerable"** - The question cannot be answered with the given database schema. This includes:
   - Questions requiring information not available in the database
   - Questions that are completely out of scope for SQL operations
   - Questions requiring functionality outside of SQL operations
   - Questions that are general conversation or small talk

# Input
- Database Schema
- Question

# Output Format
Respond with a single JSON object:
{{
  "reasoning": "<reasoning behind your decision>",
  "answer": "<one of the three categories: "answerable", "ambiguous", or "unanswerable">"
}}