Do you have a python script demo corresponding to your "High-level architecture flow diagram"?

[wonders7796]

You mentioned in your readme "The orchestrator then saves the conversation, updating the Agents' Conversation history, before delivering the response back to the user. " , how to implement it, is there a corresponding python script?

[swartz-k]

REF: python/src/multi_agent_orchestrator/orchestrator.py#L181

route_request serves as the entry point for handling user requests, the main steps being to find a suitable agent through classify_request and then call agent_process_request.
In the agent_process_request function, it is mainly divided into

1. dispatch_to_agent: Organize data, retrieve session history, call agent
2. create_metadata: Organize metadata data
3. save_message: Store ConversationMessage in storage

You might be referring to 3?