auth-helpers VS quivr

I have encountered a similar issue, but I haven't fully found a solution yet. Currently, I'm using the getSession function throughout my server components. Yesterday, I tried running a middleware to refresh the user's session, based on this example: https://github.com/supabase/auth-helpers/blob/main/examples/nextjs/middleware.tsx. This caught my attention because the comment mentioned that the middleware is necessary for any Server Component route using createServerComponentClient. Right now, I'm testing this approach, and so far, I haven't been logged out since yesterday. Do you have that in your application? The main difference between getSession and getUser in supabase is that getSession reads the cookies and is very fast (around 1-2 ms), while getUser makes a call to the supabase API with the access token to retrieve the user and I would assume they also verify the JWT on their end.

F.ex. Have a look how supabase-auth-helper-sveltekit works. This is the "SvelteKit way" to approach authentication. It also works perfectly with invalidate to reload or invalidate data that is connected to the current user session. As pointed out already, there is no reason you should not be able to access your JWT.

We have updated the Next.js Auth Helpers package to make it available across the client and server of the App Router. They also now implement server-side auth by default with PKCE - meaning the entire auth flow is now possible server-side.

Example

seems like known issue https://github.com/supabase/auth-helpers/issues/497

auth helper for sveltekit: https://github.com/supabase/auth-helpers/tree/main/packages/sveltekit

I do think checking auth in a server component would likely be better for your use here. Supabase has a nice example here: https://github.com/supabase/auth-helpers/tree/main/examples/nextjs-server-components They have a simple case with Middleware, but you wouldn't need to do it that way. Can use "redirect(/login)" from the server component.

3. Quivr | GitHub | tutorial

I used Ollama (with Mistral 7B) and Quivr to get a local RAG up and running and it works fine, but was surprised to find there are no easy user-friendly ways to do it. Most other local LLM UIs don't implement this use case (I looked here), even though it is one of the most useful local LLM use-cases I can think of: search and summarize information from sensitive / confidential documents.

In my opinion LLM based document search tools such as OSS Quivr may be better suited for documentation search for startups.

A highly customed Quivr with one of the 'Open Source LLMs' may provides great 'semantic search' for product documentation.

https://github.com/StanGirard/quivr

Hey thanks for creating this, will try later if i have time. Meanwhile, do you try some of other second brain app such as this, and how was the comparison? The one i mentioned was trending on github so i think its decent (been playing with it since last week or so, also). But i already starred your repo so i can come back later.