# Context preservation across new chats, similar to "Gems" in Gemini

**URL:** <https://replit.discourse.group/t/context-preservation-across-new-chats-similar-to-gems-in-gemini/4315>\
**Category:** Feature Requests\
**Tags:** agent\
**Created:** [April 16, 2025, 1:18pm UTC](https://replit.discourse.group/t/context-preservation-across-new-chats-similar-to-gems-in-gemini/4315 "2025-04-16T13:18:40Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![jorge130](https://yyz2.discourse-cdn.com/flex010/user_avatar/replit.discourse.group/jorge130/32/1180_2.png) [@jorge130](https://replit.discourse.group/u/jorge130)\
**Post date:** [April 16, 2025, 1:18pm UTC](https://replit.discourse.group/t/context-preservation-across-new-chats-similar-to-gems-in-gemini/4315/1 "2025-04-16T13:18:40Z")

</div>

Hello,

As the application I’m working on has grown in complexity, I’ve realized that it would be great to have a feature where we can save context instructions for new chat prompts, similar to the process @CulturaFutura described in this post:

> [@Learning to Make Changes Across Multiple Conversations](https://replit.discourse.group/t/learning-to-make-changes-across-multiple-conversations/1208):
>
> Hey folks, I’m new to these forums but I’ve been impressed by all the creative things others have been sharing, so I thought I’d break out of my “lurker” status and actually try to contribute something of my own! Something I thought might be worth sharing is a method for reducing errors that I stumbled onto yesterday. The TLDR is that my process now involves having the Agent create a development changelog that it a) reads thoroughly before beginning a new conversation window, and b) updates ev…

The idea would be to be able to save these instructions into “containers” that we could reference when starting a new task, similar to what Gemini is doing with its “Gems” feature ([Gems - build custom AI experts from Gemini](https://gemini.google/overview/gems/?hl=en)).

We would then access them from the “New Chat” button, which could look something sort of like this:

 ![jreyna_replit_feature_request](https://canada1.discourse-cdn.com/flex010/uploads/replit/original/2X/5/5d968014544122830be448118e80781d3cf6023f.png)

This would help increase productivity since we’d be saving time repeating instructions in prompts.

I’ve started using a README.md file that contains the broader context, but I still have to type an instruction (“Please review the README.md file at the root of the project to understand our established guidelines and preferences before assisting with this task.”) after I start a new chat. And then, I typically have to instruct it further with other markdown files I’ve used to document specific features.

It would be great if we had a tool that helped us achieve that much faster!

Jorge