projects
-
local-rag
A local RAG (Retrieval-Augmented Generation) library that plugs into the Vercel AI SDK for context retrieval. Persists to a local JSON file by default. For small apps, enable these options: - **FileStore** – Async I/O,…
-
chattui
Terminal chatbot using: - **Vercel AI SDK** – , - **Ollama** – chat model (Ollama Cloud) - **OpenRouter** – embedding model - **Ink** – terminal UI (React for CLI) - **local-rag** – RAG with shared store from Loads and…
-
mobile-responsive-sonner
A tiny wrapper around sonner that automatically picks the best toast position based on device type. Instead of hardcoding toast positions everywhere, this picks sensible defaults for you: - Mobile → top-center - Desktop…
-
persisting-network-calls-nextjs
I was building an AI chatbot, I know I know the classic AI chatbot ... whatever. Well I have been building one, using vercel's AI SDK and I came across an interesting problem. Whenever I refresh the page, the chatbot…