Skip to content

The Open Standard for Generative UI

2.3 KB
Raw Blame

This is an OpenUI Self Hosted Chat project bootstrapped with openui-cli.

OpenUI: guide for coding agents

This project is built with OpenUI. Before changing it, read the maintained OpenUI skill for coding agents published at thesysdev/skills.

Setup

Create .env.local with your OpenAI credentials:

OPENAI_API_KEY=...
# Optional:
OPENAI_MODEL=gpt-5.2

Getting Started

First, run the development server:

pnpm dev

Open http://localhost:3000 with your browser to see the result.

You can start editing the page by modifying src/app/api/chat/route.ts and improving your agent by adding system prompts or tools. A LangGraph scaffold puts the implementation in src/agent.ts instead.

The generated app includes a get_weather tool example. Ask “What’s the weather in Berlin?” to exercise its native tool loop.

Deploy

From the project directory, deploy a preview with the pinned OpenUI CLI:

pnpm run deploy
pnpm run deploy -- --prod

The command deploys to Vercel. Allowlisted keys from .env / .env.local (including OPENAI_API_KEY) are passed to that deployment unless you use --skip-env. Persist them on the Vercel project for later deploys.

Framework deployments

The Vercel AI SDK scaffold runs its backend inside the Next.js API route, so the frontend and backend can be deployed together as one Next.js project.

Conversation storage

This starter does not configure durable conversation storage. AgentInterface keeps messages in memory for the current page session and sends that history to /api/chat; refreshing the page loses it. To persist conversations, pass a storage implementation to AgentInterface and back it with your own database. Add a LangGraph checkpointer separately only for graph-specific durable state.

Learn More

To learn more about OpenUI, take a look at the following resources:

LangGraph is the default backend and uses Chat Completions with your provider.