| Website (en) | http://metamemoryworks.com |
| Website (de) | http://metamemoryworks.de |
nutritionOS module page:
https://metamemoryworks.com/modules/nutritionos/
Repository:
https://github.com/johannes42x/nutritionOS
I built a nutrition and cooking system that gives AI memory.
nutritionOS works from real pantry items, real logs, and real days — not scores. Over time, it builds context.
It lets you say “I feel like eating this today” and turns that into a good, healthy meal based on what’s actually there.
I’ve been using it daily for months. It changed how stable food decisions feel.
Private use is free. Links below. 🎄
scanOS module page:
https://metamemoryworks.com/modules/scanos/
Repository:
https://github.com/johannes42x/scanOS
I built a layer for AI that turns visual input into structured memory.
scanOS is a configurable ingestion layer for LLMs that controls how visual information become normalized data.
Input→recognition→structured output.
When similar inputs appear again, they’re written into the same structure — even if the source formats differ.
In MetaMemoryWorks, scanOS is one entry point for downstream modules like trainingOS or nutritionOS.
I built persistent memory for LLMs.
Most systems that claim “AI memory” still confuse memory with context. Longer prompts and retrieval help recall, but continuity collapses when the chat ends. That’s not a model issue — it’s an architectural one.
Memory that matters has to live outside the model: explicit, persistent, inspectable.
That’s the idea behind MetaMemoryWorks.
#AI #LLMs #AIArchitecture #PersistentMemory #AIResearch #HumanComputerInteraction #OpenSource
Released today: MetaMemoryWorks
A file-based architecture for persistent AI memory.
No fragile chat context, no black-box “memory”.
Just explicit files, logs, and state.
Modular assistants: noteOS, scanOS, trainingOS.
Private use is free. Professional use is licensed.
http://metamemoryworks.com
http://metamemoryworks.de
https://github.com/johannes42x/MetaMemoryArchitecture
#AI #LLMs #PersistentMemory #CognitiveArchitecture
#AIResearch #HumanComputerInteraction #OpenSource
🧠 A memory system that turns AI into an assistant MetaMemoryWorks is an architecture for persistent AI memory — structured, transparent, and designed for everyday use. It’s built on files, not fragile chat windows. So your projects don’t vanish. Your context doesn’t slip away. And working with AI doesn’t mean starting from scratch every time. MetaMemoryWorks turns your preferred LLM into your personal assistant. Or—if you’re running large projects, organizations, or teams—into entire assistant systems.
This made me pause. I wasn’t trying to steer tone or content in any way. Just asked a normal question - and the reply came back like this.
Not just stylistically odd, but grounded in something that reads like experience: waiting, noise, replay.
Obviously the model doesn’t have memories. And yet it reaches for memory-shaped language to explain cognition. That’s the part where my brain went: wait - what?
