Define Pydantic Models for Policy & Compliance

Begin by defining several Pydantic models that will represent the structured data your agent will extract and generate. Create a `TaxExemptionPolicy` model with fields like `entity_type`, `exemption_type`, `scope`, `conditions`, and `valid_until`. Also, define a `ComplianceReport` model that includes `compliance_status`, `recommendations`, and `relevant_clauses`. These models will guide your agent's output and validation logic.

```python
from pydantic import BaseModel, Field, conlist
from typing import Optional

class TaxExemptionPolicy(BaseModel):
    entity_type: str = Field(..., description='Type of entity eligible for exemption (e.g., foreign_cloud_provider)')
    exemption_type: str = Field(..., description='Type of tax exemption (e.g., income_tax, import_duty)')
    scope: str = Field(..., description='Specific scope of the exemption (e.g., services_sold_outside_india)')
    conditions: conlist(str, min_length=1) = Field(..., description='List of conditions that must be met for the exemption')
    valid_until: Optional[int] = Field(None, description='Year until which the exemption is valid, if applicable')

class ComplianceReport(BaseModel):
    compliance_status: str = Field(..., description='Overall compliance status (e.g., Compliant, Non-Compliant, Conditional)')
    recommendations: conlist(str, min_length=1) = Field(..., description='Actionable recommendations for ensuring compliance')
    relevant_clauses: conlist(str, min_length=1) = Field(..., description='List of relevant policy clauses or sections')

print(TaxExemptionPolicy.model_json_schema())
print(ComplianceReport.model_json_schema())
```