Implement chat.py functionality using LangGraph

[TanGentleman]

The Chatbot class was made to manage state. Now we're going to rewrite it using LangGraph concepts.

Roadmap:

• Set a standard for the documentation of chat commands and their effects. This ended up getting messy in the original implementation.
• Move deterministic flows into chains, avoid importing many functions to the file with nodes

Notes:

• Use LangChain's serialization while testing, it's handy. https://python.langchain.com/docs/how_to/serialization/#saving-objects
• Be very deliberate when exposing a key to the state.

[TanGentleman]

Design Patterns

There are usually three different actions you can take with a human-in-the-loop workflow:

• Approve or Reject: Pause the graph before a critical step, like an API call, to review and approve the action. If the action is rejected, you can prevent the graph from executing the step and potentially take an alternative action. This pattern often involves routing the graph based on the human’s input.
• Edit Graph State: Pause the graph to review and edit the graph state. This is useful for correcting mistakes or updating the state with additional information. This pattern often involves updating the state with the human’s input.
• Get Input: Explicitly request human input at a specific step in the graph. This is useful for collecting extra information or context to guide the agent’s decision-making process or for supporting multi-turn conversations.