feat: add Cerebras as an AI provider in backend and UI

[krishsharma-code]

Overview

This PR implements the backend connector and frontend UI configuration for adding Cerebras AI as a natively supported provider. Cerebras is OpenAI-compatible, so it correctly routes through the existing OpenAI wrapper factories.

Changes Made

• Backend (internal/aiconnectors & internal/ai/langchain):
  • Added ProviderCerebras constant and mapped it to createOpenAIModel().
  • Added canonical base URL (https://api.cerebras.ai/v1) to defaults.
  • Set default fallback model to llama3.1-8b and token limit cap to 8000.
• Frontend (ui/):
  • Updated AIProviders.tsx to add Cerebras in the UI selectable list with placeholders for API keys and auto-populated base URLs.

Related Issue

This PR fulfills the backend implementation required for HexmosTech/git-lrc#79 under GSSoC.

[shrsv]

Can you share a video walkthrough of how the UI looks like @krishsharma-code (so we can give UI flow feedback if any)?

Also if flow overall OK - we also have workflows to run (security, organization, etc) as well - which we will do once flow seems OK. We will also do a manual review once it runs.

[krishsharma-code]

Uploading Screen Recording 2026-05-27 213028.mp4…

Hey! I have successfully tested the Cerebras provider locally using Docker. The UI correctly auto-fills the default model (llama3.1-8b), and the backend API validation is correctly triggering and catching invalid test keys. Attached the screen recording for reference! I successfully spun up the entire LiveReview stack locally via Docker and tested my Cerebras AI provider implementation. The default model auto-fills perfectly and backend validation is working exactly as expected.

[LinceMathew]

hi @krishsharma-code, Thanks for the video demo, btw. Could you please try with a valid API key? Basically we need to make sure the AI review is also working with this Cerebras. Let us know if you have any issues or doubts.