Opt-in user memory example for the LangGraph ReAct starter #11

Hi maintainers. I’m Vivek Gupta, Founder & CEO at MemCode (https://memcode.in). This LangGraph ReAct starter already illustrates MCP tools, model choices, and agent evaluation. An optional example could show how an application carries one user-approved preference into a separate conversation while keeping graph state and evaluation traces independent.

I can contribute a MemCode MCP tool recipe with explicit save and recall calls, user scoping, and a test that two users cannot see each other’s records. The starter would keep its current behavior unless developers enable the recipe. Prompts, tool traces, credentials, and full transcripts would not be sent automatically. Would a tool example or a store adapter fit the repository better?

cc @webup for your thoughts on this integration.