Add example: agent with structured output validation (Pydantic)

[rajkumar42]

Summary

Create an example that demonstrates using Pydantic models to validate and enforce structured output from an OpenSymbolicAI agent. Shows how to guarantee the agent returns data in a specific schema.

What needs to happen

• Create structured_output/ directory
• Define a Pydantic model for the expected output (e.g., ExtractedContact with name, email, phone)
• Create an agent that extracts structured data from unstructured text
• Validate the agent's output against the Pydantic model
• Handle validation errors gracefully (retry or report)
• Add README explaining the pattern

Example

from pydantic import BaseModel
from opensymbolicai import Agent

class Contact(BaseModel):
    name: str
    email: str
    phone: str | None

agent = Agent(
    goal="Extract contact info from this text",
    output_schema=Contact
)
result = agent.run(input="Call John at john@example.com or 555-1234")
print(result)  # Contact(name='John', email='john@example.com', phone='555-1234')

Acceptance criteria

• Output is validated against a Pydantic model
• Invalid outputs are caught and handled
• README explains when and why to use structured output