MetaMemoryWorks turns AI into an agent for your everyday life. You simply use one or more of the available modules. trainingOS, for example, turns your AI into a personal training assistant. It doesn’t just help you design your training routines. trainingOS remembers what you actually do in your workouts: repetitions, weights, distances, times… whatever matters to you.
You keep using your familiar LLM. Instead of fleeting chats, you work with files, modules, and a memory that stays. Free for private, non-commercial use and fully adaptable to your needs 🙂
On this page, you’ll learn how to get started in practical terms. No prior knowledge required.
0. What would I even use an AI assistant for?
Imagine you have a device in your pocket that can think and answers every question exactly the way you want. Or reacts to any instruction exactly the way you’ve defined. AI doesn’t think. It imitates thinking. And it does so well that, in everyday life, the difference often doesn’t matter anymore.
What does make a difference in daily life is an assistant: on your smartphone or on your screen, available around the clock.
MetaMemoryWorks turns your AI into a coach, a nutrition advisor, a sleep expert, an archivist, a digitization tool, an application coach, a negotiation trainer, a learning and working environment, a playground for synthetic personas – or even a synthetic theater machine.
This works because the architecture gives AI a memory it can restart from every time, exactly the way you need it. And because it remembers how you last used it – when you save.
Instead of starting from scratch every time, you build something step by step that grows with you.
And if you need an assistant that doesn’t exist yet? As long as you’re using it privately and non-commercially, the architecture is flexible enough. You just build it yourself 😉
1. What you need (really not much)
✔️ Access to an AI
You need an account with an AI provider: ChatGPT, Claude, Perplexity, Copilot… whatever you prefer. The only important thing is that you can write text and upload files. And then you’ll need a module. You can find those here.
✔️ A place for your files
Store the files on your device in a way that you can easily find them again. A proven approach is one folder per module, with subfolders for logs and other files that grow over time.
2. The core principle (really 30 seconds)
MetaMemoryWorks always follows the same simple flow.
1️⃣ Download the module files
You download the files for a module.
Usually, you’ll get a .zip file. It contains the files the module consists of, typically:
- an engine
- a log
- a start guide
- legal information
This is your starting point.
2️⃣ Load the module into your LLM
Example: ChatGPT
- Create a new project on the left (click “New project”) and name it something like “trainingOS”.
- Open the project.
- Above the input field, you’ll find the file upload.
- Upload all module files. You don’t need to upload terms, disclaimers, or licenses – just make sure you’ve read them.
- That’s it. The module is now active. You can start working with your assistant in the chat. Feel free to ask it first what it can do.
💡 Note:
Without ChatGPT Plus, you can upload five files per project.
That’s usually enough.
Each module description states how many files are required.
3️⃣ Use it: work, save, remember
How exactly you use a module depends on the module itself. You’ll find an overview on the modules page.
In general:
- What a module can do and how to use it is explained in the start guide inside the module folder.
- You simply work via the chat window. Tell the assistant what you want to do.
- Confused? Ask. “Huh?” actually works surprisingly often – but more specific questions will get you there faster 🙂
- When you’ve logged something (the exact instructions are in the module’s start guide), the assistant will give you an updated file.
- Save that file. Important: If, for example, an updated log is smaller than the existing one (check the file size!), the AI made a mistake. Just tell the assistant and ask it to append the entry properly.
- Upload the updated file back into the project context and remove the old one.
These files are the memory of your AI.
Important:
All files belong to you.
They are processed only by you and your AI provider –
never by MetaMemoryWorks itself.
3. Everyday assistants
Now it’s time to try things out: explore what the modules can do and how they can make your daily life easier and lighter.
You’re holding a powerful tool in your hands – one that can work with anything you can express in language.
And because you’ve just set up an assistant for this tool, you can now tell it what you want to do with it. It will remember.
Have fun building your personal assistants!
MetaMemoryWorks – do EVERYTHING!