feat: structured output and tool calling examples for Crusoe Managed Inference

[Sakshi3027]

What this adds

Four working demos of structured output and tool calling with ChatCrusoe.

Demo	What it shows
Structured summary	Pydantic-validated fields from free-form LLM output
Code review	Typed verdict, score, and issue lists
Entity extraction	Companies, people, technologies, locations from text
Tool calling	Bind tools to the model and parse tool call arguments

Why it's useful

Structured output and tool calling are core patterns for production LLM applications — they make model responses reliable and machine-parseable. The langchain-crusoe README documents that these features are supported, but there are no solution examples showing how to use them end to end. This fills that gap.

Testing

All 4 demos run successfully. Tested locally with Groq as a drop-in replacement.

To run on Crusoe:
export CRUSOE_API_KEY="your-api-key"
python structured_output.py

Related contributions

• feat: LangGraph multi-node research agent on Crusoe Managed Inference #55 — LangGraph multi-node agent on Crusoe
• feat: MLflow experiment tracking for Crusoe Managed Inference #56 — MLflow experiment tracking on Crusoe
• feat: RAG pipeline with Qdrant and sentence-transformers on Crusoe Managed Inference #57 — RAG pipeline with Qdrant on Crusoe