JevMade Sign in
← Back to guides

JevMade field notes / Architecture and workflow guide

How to safely store and share saved instructions for AI assistants

This guide explains how to build a private library of saved instructions for an AI assistant. It covers how to keep different users separated, suggest useful instructions, and block risky files from being added.

Original by KitzeAgent workflows

Listen to this guide

JevMade’s plain-English explanation

0:00 /

Our summary

This software acts as a private library for saved instructions that tell an AI assistant how to do specific tasks. People use it to store these instructions safely and control exactly who is allowed to read them.

When an AI assistant needs help, the library uses Jev, an AI tool that chooses from options rather than writing an answer, to suggest the best instructions. These suggestions are only hints, and the assistant must decide whether to use them.

This system is useful for people managing multiple users because it strictly separates what each person can access. In the authors' tests, it blocked imported files that were too large, broken, or contained private access codes.

Key takeaways

  1. Set up the library using separate random administrator codes so it will not overwrite existing secret files.
  2. Treat the suggested instructions as helpful hints rather than commands, and never share them between different users.
  3. The system blocks imported files containing private access codes, broken data, or oversized text.

The authors reported how this software behaves and measured its limits.

Repository README and documentation

Read the original guide Opens the author’s site in a new tab.

Keep this for later

Sign in to bookmark experiments, guides and videos, and keep notes only you can see.

Continue to sign in

We’ll bring you back to this listing.