Relay Chat Demo

A simple Next.js App Router scaffold for a chat app built with the Vercel AI SDK and the direct OpenAI provider.

What is included

• A streaming chat interface powered by useChat
• A ToolLoopAgent with three mocked server-side tools
• Typed UI messages via InferAgentUIMessage
• An API route that streams agent responses with createAgentUIStreamResponse

Setup

1. Install dependencies:

npm install

2. Create a local env file from the example:

cp .env.example .env.local

3. Add your OpenAI API key to .env.local:

OPENAI_API_KEY=your_key_here

4. Optional: add PromptLayer tracing credentials if you want OpenTelemetry traces exported to PromptLayer:

PROMPTLAYER_API_KEY=your_promptlayer_key_here
OTEL_SERVICE_NAME=vercel-chat-app

5. Start the app:

npm run dev

Open http://localhost:3000.

Model configuration

The scaffold uses the direct OpenAI provider from @ai-sdk/openai and defaults to gpt-5, matching the current example on the AI SDK OpenAI provider docs. You can override that with:

OPENAI_MODEL=your_preferred_model

PromptLayer tracing

This app can export Vercel AI SDK traces to PromptLayer using OpenTelemetry.

• PROMPTLAYER_API_KEY enables OTLP trace export to PromptLayer
• OTEL_SERVICE_NAME optionally overrides the service name shown in traces

If PROMPTLAYER_API_KEY is not set, the app still runs normally and skips PromptLayer export.

Useful scripts

• npm run dev
• npm run lint
• npm run typecheck