Quick overview (60 seconds)
What is MetaMemoryWorks?
An architecture made of files and modules that gives your preferred LLM a reliable, structured memory – instead of starting from scratch over and over again.
Why would I need this?
Because chats are ephemeral.
What you explain today is often gone tomorrow.
With files, your workflow stays stable – and AI can actually build on what came before.
Is this an app or a service?
No.
You don’t install anything.
You simply work with files and an LLM of your choice – the same way you already do.
Do I need programming skills?
No.
MetaMemoryWorks is explicitly designed to work without technical prior knowledge.
How do I get started?
You choose a module, download a few files, and work normally in the chat.
A step-by-step guide can be found on the “Getting Started” page.
Do I have to commit to anything?
No.
You can use modules individually, add more at any time, or combine them.
Each module works in its own set of files.
What does MetaMemoryWorks cost?
Using the architecture and modules is free for private, non-commercial use.
If you want to save time during setup, you can optionally purchase presets.
What happens to my data?
MetaMemoryWorks does not store any user data.
Your files belong to you and stay with you – and in the LLM you choose to use.
MetaMemoryWorks never sees them.
A. Basic understanding
What makes MetaMemoryWorks different from typical AI workflows?
MetaMemoryWorks does not rely on chat histories, but on clear, reproducible files.
This keeps your work traceable, reusable – and prevents it from getting lost.
Can any AI work with this?
Yes, in principle. MetaMemoryWorks has been tested with ChatGPT.
The architecture itself is general and should also work with other LLMs.
Please be cautious with preset purchases in the early days if you’re using a different LLM.
B. Usage & getting started
How do I load modules into ChatGPT or other LLMs?
In ChatGPT: via a Project or a Custom GPT where you upload the module files.
Other LLMs should work similarly, as long as they accept text and file uploads.
How do I store logs correctly?
Simply as files in the corresponding module folder.
The only rule is: never overwrite, always append.
What if a log wasn’t written correctly?
That happens sometimes.
Just ask the LLM again to append the entry to the existing log file.
What should my folder structure look like?
Whatever feels clear and manageable to you.
A common pattern is to keep logs in their own subfolder.
If you keep older logs, you can immediately see whether a log actually grew –
or whether the LLM may have made a mistake.
C. Modules
What is an “OS module”?
A bundle of files that defines:
- how an LLM should behave
- how it stores information
- and how those files work together
How do presets work, and do I need them?
Presets are convenient, but optional.
They simply save you setup time.
What’s the difference between base packages and presets?
Base packages = fully functional.
Presets = convenience.
What’s coming next?
New modules are released regularly – nutrition, sleep, learning, careers, organizations, creativity, and more.
D. Data & privacy
Important: You can also find information about data protection in the privacy policy.
Who has access to my logs?
Only you – and the LLM platform you’re using.
Where are my files stored – locally or in the AI?
With you: on your device or in your cloud storage.
And additionally in the LLM where you upload them yourself.
E. License & usage
Is private use really free?
Yes.
Private, non-commercial use by individuals is completely free.
Do I need a license for professional or institutional use?
Yes.
Companies, organizations, public administrations, and NGOs require a commercial license.
What about universities and research?
Institutional research generally requires a license.
A license is not required if the use is:
- purely analytical or evaluative,
- not integrated into operational systems, services, or products,
- not used for behavioral control, surveillance, or automated decision-making,
- and strictly limited to research or publication purposes.
If in doubt, please just ask.
What is explicitly prohibited under the license?
Military, repressive, or manipulative applications,
as well as anything that harms people.
Why is there an anti-abuse clause?
Because technology always comes with responsibility.
F. Technology & LLM behavior
Why is persistent memory so difficult for AI?
Because chat histories are ephemeral.
Anything that isn’t explicitly stored is usually lost by the next conversation.
Why do files work better than chat histories?
Because they remain stable, reusable, and traceable –
and because you control them yourself instead of relying on chat logs.
How do I make sure logs are truly “append-only”?
The simplest way is to check the file size –
or to explicitly instruct the LLM to only append.
Can MetaMemoryWorks be used offline?
The files: yes.
The AI: of course not.
G. Miscellaneous
Why does MetaMemoryWorks feel so calm and structured?
Because you never start from “zero” again.
What’s been clarified, decided, or written down stays there –
and everything has its place.
Is there a community or ways to exchange ideas?
Not yet.
Whether and in what form a community emerges – Discord, subreddit, forum, or something else –
will be decided after the release.
Who is Kath?
Kath is CMO and press spokesperson of MetaMemoryWorks.
And a student at the Diederich-Heßling-Gymnasium in Fulda.
If you can’t find something or feel something is missing,
the module start guides usually go into more detail.
MetaMemoryWorks – do EVERYTHING