• https://langchain-ai.github.io/langgraph/tutorials/langgraph-platform/local-server/

Takeaways:

• Pass configs into tools using runnable_config. Use type annotation InjectedToolArg
• Use env variables to configure tools whenever relevant to different deployments
• Consider runnable configs seriously as this is by design of LangGraph
• Any graph can be used with LangGraph Studio
• Next steps: still to play with deterministic tool nodes