Remotely Stored Repository Knowledge Graphs to not regenerate individually

[Den-MM]

It rather sounds like an enterprise feature or something, but caching and storing existing graphs remotely that are mapped to frequently generated knowledge graphs for certain repositories would avoid users burning a lot of tokens when generating graphs locally and individually.
That's a very broad AI issue in general, though, that people would regenerate a lot of ai resources for themselves although with all this material, knowledge, etc., could be stored centrally somewhere for quick reuse instead of full regenerations.

[Lum1104]

https://understand-anything.com/demo/

Demo page is using something similar, it receive a remote url of the graph and render it.

[Den-MM]

https://understand-anything.com/demo/

Demo page is using something similar, it receive a remote url of the graph and render it.

Exactly, but see, what I thought of was that many gitHub repositories are ran with /understand currently with your tool but we have no general server to be storing those graphs that have already been generated.
But it's not something that should be your responsibility since it would need database maintenance etc.

What you probably meant with your answer was that remote URLs of the graphs exist currently and the graph can render with it, but what I meant is collecting graphs literally on a server for reuse. But I understand this goes more into a fully featured remote service then (like youtube storing yt videos), rather not into a direction of providing claude /ai plugins of what your repo rather is for.

[MrSuddenJoy]

@Den-MM this is achievable, but with external tooling (import as dependency)