Injecting and retaining large contexts to be used by default

Does anyone know of a method to inject and store/retain large context (5000+ token) without having to pass it in each prompt chain? Something similar perhaps to the “memory” function seen in the retrieval plugin use case, or potentially achieved through fine-tuning to teach the model to respond only according to a large default context?

This would enable you to save the contextual state of your gpt powered app, so users would only receive responses according to that context, without having to pass it each time in the prompt chain.

I’m thinking this must be a common use case