Ongoing development of OpenMemory?

[ndjones]

Hi @nullure - OpenMemory looks well conceived and has some interesting and useful approaches to this agent memory space. Activity looks like it has dropped off a bit since last year, especially over the last 6 months. Is there anything going on in the background out of tree, or is the state of the repo here accurately capturing all activity? It looks like you had started a rewrite in May that has stalled. Just wondering if there is an ongoing intention to develop further, as it looks interesting.

[acester822]

I went down a major rabbit hole with this awhile back and ended up forking where openmemory was at and splicing in improvements along the way, take a look!

https://github.com/acester822/FTR10-Engram

I still have some tweaking to do, but it works well. If time allows after I get the main engine humming I intend to add true benchmarking as well because I am seeing a flurry of apps like mine that claim to do the same thing or close to it, and 95% of them are hallucinated AI claims and the apps are nothing more than RAG tools that barely work.

[nullure]

We are currently working on a rewrite, testing it across multiple tools and systems. It is expected to be released around the first anniversary of OpenMemory or sooner.

[ndjones]

sounds promising thanks @nullure - i gained some insights from analysing the current code base, happy to share if useful. good luck!