Skip to content

Conversation

@franciscojavierarceo
Copy link
Collaborator

@franciscojavierarceo franciscojavierarceo commented Aug 27, 2025

What does this PR do?

This PR adds:

  1. VectorDB creation in the Agent Creation Configuration
  2. File Upload using the RagTool
  3. A more welcoming onboarding experience for creating an agent
  4. A handful of other implementation details to get this to work

I used Claude to help me write a meaningful share of the code but I made lots of revisions. I also used AI for most of the text copy. I vetted the code but I am not a frontend expert so some areas I am still learning (e.g., uploading larger PDFs required me learning a bit more about this space, I think the solution is the right one).

The important part is I think we're at a good starting point for an MVP based on the existing APIs and not having a supported UI is probably preventing the community from further growth, so I feel good with landing this.

Screen Shots with Demo

Clean starting point when no agent sessions configured

Screenshot 2025-08-27 at 4 45 26 PM

Creating a Vector DB (section in Agent Creation Config)

Screenshot 2025-08-27 at 4 47 23 PM

Creating a Vector DB (actual modal)

Screenshot 2025-08-27 at 4 48 08 PM

Uploading document notification

Screenshot 2025-08-27 at 4 41 41 PM

Successfully uploaded document notification

Screenshot 2025-08-27 at 4 42 45 PM

Functional RAG Example

Screenshot 2025-08-27 at 4 39 07 PM

Closes #2307

Test Plan

Unit tests added and manually tested edge cases.

@meta-cla meta-cla bot added the CLA Signed This label is managed by the Meta Open Source bot. label Aug 27, 2025
Copy link
Collaborator

@leseb leseb left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

No UI expert here, thanks for sharing the screenshots, this looks nice. A step in the right direction for sure. We can iterate if needed as users start consuming it. Thanks!

Rubber stamp!

ps: you really need to do something about those tabs… I can hear your browser screaming xD

@franciscojavierarceo franciscojavierarceo merged commit 75fad44 into llamastack:main Aug 28, 2025
5 checks passed
cdoern added a commit to cdoern/llama-stack that referenced this pull request Sep 5, 2025
when using the providers.d method of installation users could hand craft their AdapterSpec's to use overlapping code meaning one repo could contain an inline
and remote impl. Currently installing a provider via module does not allow for that as each repo is only allowed to have one `get_provider_spec` method with one Spec returned

add an optional way for `get_provider_spec` to return a list of `ProviderSpec` where each can be either an inline or remote impl.

Note: the `adapter_type` in `get_provider_spec` MUST match the `provider_type` in the build/run yaml for this to work.

resolves llamastack#3266

Signed-off-by: Charlie Doern <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

CLA Signed This label is managed by the Meta Open Source bot.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Developer-facing UI for chat completions

2 participants