Looks like it is a crowded area now - my angle is to start with theory of what is important in a system like that, from first principles (like agent limited context, statelessness, use goals etc). Currently I use it to develop that theory - and you can read it at: https://zby.github.io/commonplace/. I also use it to keep an index of similar systems (that is systems with agent operated memory): https://zby.github.io/commonplace/agent-memory-systems/

The github repo is at: https://github.com/zby/commonplace . Work in progress.