If you are interested you can try it out at markbase.cloud (disclaimer and all that). I am not charging for it.
It's hedging a bet at this point, but that's why people say there's no moat. If the tools are properly used + maintained, there should be no reason we can't use a new provider even next week (maybe with a little tweaking).
this is how it works: https://help.markbase.cloud/humans/collections/overview
https://arxiv.org/abs/2602.11988
For what it's worth, if you were considering building context out.
Unless you work in some obscure domain, chances are that any general "knowledge" Claude has "learned" is already public data somewhere.
If you don't believe me, launch Codex and immediately start working on the same project (s). You might discover that all the knowledge accumulated means almost nothing.
This isn't something that is public knowledge, in the sense that you mean it.
Just earlier today it asked me if I wanted to create a jira ticket for something I asked it about doing. My prompt mentioned nothing about jira.
If you use Claude Code, you might want to take a look at the "auto memories" files that it creates. See "/memory" for some more information.
Where is the knowledge stored?
All of my knowledge typically gets stored in plans outside of the agent?
And each agent window gets archived regularly, anyways.