Justyn Shull

> editor_ui.tsx:146 Error running command Function searchCommand is not available in client At least this part is because I had `env: server` on searchCommand. I removed it and it seems...

I added a `system.invokeFunctionOnSystem` syscall that just always uses `proxySyscall`. I'll do some more testing with it and create a PR to the main silverbullet repo if you think that'd...

I changed the results page to look like this: I'm also moving all of the search logic out of the pageNamespace readFile function and instead triggering it on an `editor:pageLoad`...

@zefhemel can you give it another try when you get a chance? I had to temporarily add a sleep in the loop to test with, but I think the experience...

Hmm, immediately after `Generating query vector embeddings`, it calls the embeddings api, and then you should see `Searching for similar embeddings`. If there was an error calling the embeddings api,...

I just pushed an update to display an error on the results page if it can't generate the embeddings, so it should at least give feedback now.

Okay I changed it so embeddings are always generated on the server.

> I'm wondering if this is something to support. It could be done by indeed proxying LLM calls through the server. That is probably doable, except for all the streaming...

@zefhemel I came back to this issue over the last week and haven't really come up with a good solution. One idea that seemed promising was to move everything to...

I need to update the readme, but this page has the information on how to enable and use embeddings: https://ai.silverbullet.md/Configuration/Embedding%20Models/ The setting is `embeddingModels` and would look like this, including...