Skip to content
Source
Sign inSign up
TanStack/ai

examples/ts-react-chat

README.md

Welcome to your new TanStack app!

Getting Started

To run this application:

pnpm install
pnpm dev

Building For Production

To build this application for production:

pnpm build

Testing

This project uses Vitest for testing. You can run the tests with:

pnpm test

Styling

This project uses Tailwind CSS for styling.

TanStack Chat Application

An example chat application built with TanStack Start, TanStack Store, and TanStack AI (our open-source AI SDK).

.env Updates

OPENAI_API_KEY=your_openai_api_key

Trying Out Lazy Tool Discovery

This example includes three lazy tools — tools that are not sent to the LLM upfront. Instead, the LLM sees a __lazy__tool__discovery__ tool that lists their names. When the LLM needs one, it discovers it first (getting the full description and schema), then calls it normally.

The lazy tools are: compareGuitars, calculateFinancing, and searchGuitars.

Test Prompts

Compare guitars — triggers discovery of compareGuitars:

"Can you compare the Motherboard Guitar and the Racing Guitar for me?"

Financing — triggers discovery of calculateFinancing:

"How much would it cost per month if I financed the Superhero Guitar over 12 months?"

Search — triggers discovery of searchGuitars:

"Do you have any guitars with LED lights or tech features?"

Multi-discovery — triggers discovery of multiple lazy tools at once:

"I'm looking for acoustic guitars. Can you search for them, then compare the ones you find, and show me financing options for the cheapest one?"

Self-correction — the LLM may try calling a lazy tool directly without discovering it first. It will get an error telling it to discover first, then self-correct:

"Compare guitar 1 and guitar 4 right now"

What to watch for

  • The LLM calls __lazy__tool__discovery__ before using a lazy tool for the first time
  • After discovery, the tool is called directly like any normal tool
  • In multi-turn conversations, previously discovered tools are usable immediately without re-discovery
  • If the LLM skips discovery, it gets an error and self-corrects

✨ Features

AI Capabilities

  • 🤖 Powered by TanStack AI with OpenAI GPT-4o
  • 📝 Rich markdown formatting with syntax highlighting
  • 🎯 Customizable system prompts for tailored AI behavior
  • 🔄 Real-time streaming responses with Server-Sent Events
  • 🔌 Connection adapters - flexible streaming architecture
  • 🛠️ Automatic tool execution loop - tools are executed automatically by the SDK
  • 🎸 Tool/function calling with guitar recommendations

User Experience

  • 🎨 Modern UI with Tailwind CSS and Lucide icons
  • 🔍 Conversation management and history
  • 🔐 Secure API key management
  • 📋 Markdown rendering with code highlighting

Technical Features

  • 📦 Centralized state management with TanStack Store
  • 🔌 Extensible architecture for multiple AI providers
  • 🛠️ TypeScript for type safety

Architecture

Tech Stack

  • Frontend Framework: TanStack Start
  • Routing: TanStack Router
  • State Management: TanStack Store
  • Styling: Tailwind CSS
  • AI Integration: TanStack AI with OpenAI GPT-4o
  • Chat Client: @tanstack/ai-react with connection adapters
  • Streaming: Server-Sent Events via fetchServerSentEvents
  • Tool Execution: Automatic loop with ToolCallManager

Routing

This project uses TanStack Router. The initial setup is a file based router. Which means that the routes are managed as files in src/routes.

Adding A Route

To add a new route to your application just add another a new file in the ./src/routes directory.

TanStack will automatically generate the content of the route file for you.

Now that you have two routes you can use a Link component to navigate between them.

To use SPA (Single Page Application) navigation you will need to import the Link component from @tanstack/react-router.

import { Link } from '@tanstack/react-router'

Then anywhere in your JSX you can use it like so:

<Link to="/about">About</Link>

This will create a link that will navigate to the /about route.

More information on the Link component can be found in the Link documentation.

Using A Layout

In the File Based Routing setup the layout is located in src/routes/__root.tsx. Anything you add to the root route will appear in all the routes. The route content will appear in the JSX where you use the <Outlet /> component.

Here is an example layout that includes a header:

import { Outlet, createRootRoute } from '@tanstack/react-router'
import { TanStackRouterDevtools } from '@tanstack/react-router-devtools'

import { Link } from '@tanstack/react-router'

export const Route = createRootRoute({
  component: () => (
    <>
      <header>
        <nav>
          <Link to="/">Home</Link>
          <Link to="/about">About</Link>
        </nav>
      </header>
      <Outlet />
      <TanStackRouterDevtools />
    </>
  ),
})

The <TanStackRouterDevtools /> component is not required so you can remove it if you don't want it in your layout.

More information on layouts can be found in the Layouts documentation.

Data Fetching

There are multiple ways to fetch data in your application. You can use TanStack Query to fetch data from a server. But you can also use the loader functionality built into TanStack Router to load the data for a route before it's rendered.

For example:

const peopleRoute = createRoute({
  getParentRoute: () => rootRoute,
  path: '/people',
  loader: async () => {
    const response = await fetch('https://swapi.dev/api/people')
    return response.json() as Promise<{
      results: {
        name: string
      }[]
    }>
  },
  component: () => {
    const data = peopleRoute.useLoaderData()
    return (
      <ul>
        {data.results.map((person) => (
          <li key={person.name}>{person.name}</li>
        ))}
      </ul>
    )
  },
})

Loaders simplify your data fetching logic dramatically. Check out more information in the Loader documentation.

React-Query

React-Query is an excellent addition or alternative to route loading and integrating it into you application is a breeze.

First add your dependencies:

pnpm add @tanstack/react-query @tanstack/react-query-devtools

Next we'll need to create a query client and provider. We recommend putting those in main.tsx.

import { QueryClient, QueryClientProvider } from '@tanstack/react-query'

// ...

const queryClient = new QueryClient()

// ...

if (!rootElement.innerHTML) {
  const root = ReactDOM.createRoot(rootElement)

  root.render(
    <QueryClientProvider client={queryClient}>
      <RouterProvider router={router} />
    </QueryClientProvider>,
  )
}

You can also add TanStack Query Devtools to the root route (optional).

import { ReactQueryDevtools } from '@tanstack/react-query-devtools'

const rootRoute = createRootRoute({
  component: () => (
    <>
      <Outlet />
      <ReactQueryDevtools buttonPosition="top-right" />
      <TanStackRouterDevtools />
    </>
  ),
})

Now you can use useQuery to fetch your data.

import { useQuery } from '@tanstack/react-query'

import './App.css'

function App() {
  const { data } = useQuery({
    queryKey: ['people'],
    queryFn: () =>
      fetch('https://swapi.dev/api/people')
        .then((res) => res.json())
        .then((data) => data.results as { name: string }[]),
    initialData: [],
  })

  return (
    <div>
      <ul>
        {data.map((person) => (
          <li key={person.name}>{person.name}</li>
        ))}
      </ul>
    </div>
  )
}

export default App

You can find out everything you need to know on how to use React-Query in the React-Query documentation.

State Management

Another common requirement for React applications is state management. There are many options for state management in React. TanStack Store provides a great starting point for your project.

First you need to add TanStack Store as a dependency:

pnpm add @tanstack/store

Now let's create a simple counter in the src/App.tsx file as a demonstration.

import { useStore } from '@tanstack/react-store'
import { Store } from '@tanstack/store'
import './App.css'

const countStore = new Store(0)

function App() {
  const count = useStore(countStore)
  return (
    <div>
      <button onClick={() => countStore.setState((n) => n + 1)}>
        Increment - {count}
      </button>
    </div>
  )
}

export default App

One of the many nice features of TanStack Store is the ability to derive state from other state. That derived state will update when the base state updates.

Let's check this out by doubling the count using derived state.

import { useStore } from '@tanstack/react-store'
import { Store, Derived } from '@tanstack/store'
import './App.css'

const countStore = new Store(0)

const doubledStore = new Derived({
  fn: () => countStore.state * 2,
  deps: [countStore],
})
doubledStore.mount()

function App() {
  const count = useStore(countStore)
  const doubledCount = useStore(doubledStore)

  return (
    <div>
      <button onClick={() => countStore.setState((n) => n + 1)}>
        Increment - {count}
      </button>
      <div>Doubled - {doubledCount}</div>
    </div>
  )
}

export default App

We use the Derived class to create a new store that is derived from another store. The Derived class has a mount method that will start the derived store updating.

Once we've created the derived store we can use it in the App component just like we would any other store using the useStore hook.

You can find out everything you need to know on how to use TanStack Store in the TanStack Store documentation.

Demo files

Files prefixed with demo can be safely deleted. They are there to provide a starting point for you to play around with the features you've installed.

Learn More

You can learn more about all of the offerings from TanStack in the TanStack documentation.

Persistent chat (/persistent-chat)

A chat that survives a full page reload on both ends. The client writes the transcript to localStorage (via localStoragePersistence), so a reload restores the conversation instantly. The server writes the same transcript, run records, and interrupt state to SQLite with withPersistence, so it survives a server restart too — the DB lives at .data/persistent-chat.db (gitignored). The SQLite backend is src/lib/sqlite-persistence.ts, a self-contained node:sqlite store built on the @tanstack/ai-persistence core store contracts — a worked example of rolling your own adapter.

Try it: send a message, wait for the reply, then reload the page. The conversation is still there. Needs OPENAI_API_KEY in .env.

Sandboxes — GitHub issue triage (/sandboxes)

Pick a harness adapter (Claude Code, Codex, OpenCode) and a sandbox provider (Docker, local process, Vercel, Daytona), paste a GitHub issue URL, and the agent clones that repo into a sandbox, investigates read-only, and reports whether the bug is still relevant and its root cause — streaming tool calls and file activity live.

Providers

  • Docker — needs a running Docker daemon. The agent runs in a node:22 container; the harness CLI is installed on first create. Heavy investigations can OOM the Docker VM (you'll see exited with code 137) — raise Docker Desktop's memory (Settings → Resources) if so.
  • Local process — runs the chosen CLI directly on your host (no isolation; dev only). On Windows the agent runs through git-bash/WSL sh (auto-located from git on PATH; override with TANSTACK_SANDBOX_SH), since commands use POSIX quoting — install Git for Windows, or set TANSTACK_SANDBOX_SH.
  • Vercel — VERCEL_OIDC_TOKEN (run vercel env pull), or VERCEL_TOKEN
    • VERCEL_TEAM_ID + VERCEL_PROJECT_ID. (Token alone falls back to OIDC and fails.) OIDC tokens are short-lived — re-pull when they expire.
  • Daytona — DAYTONA_API_KEY.

Harness keys

Set the chosen harness's key in .env.local (read by the dev server): ANTHROPIC_API_KEY (Claude Code / OpenCode), CODEX_API_KEY (Codex). For sandboxed providers (Docker/Vercel/Daytona) the key is injected into the sandbox; for local process the host CLI uses your own host auth (the env key, or a claude login). Optional GITHUB_TOKEN for private repos / higher rate limits.

Use your Claude subscription instead of the API (no API billing)

When running Claude Code on the Local process provider, tick “use my subscription” to bill your logged-in Claude Pro/Max account instead of the Anthropic API. It works by scrubbing ANTHROPIC_API_KEY from the spawned claude's environment so the CLI falls back to your claude login credentials:

import { localProcessSandbox } from '@tanstack/ai-sandbox-local-process'

// Drop env vars before spawning so the host CLI uses its own stored auth.
localProcessSandbox({ scrubEnv: ['ANTHROPIC_API_KEY'] })

Requires claude login (a Pro/Max subscription) on the host. Only local-process can do this — sandboxed providers have no host login, so they always use an API key. (The -p headless flag is unrelated to billing; it's just streaming mode.)

Tool bridge & code mode

The triage run passes two chat() tools to the harness to exercise the sandbox tool bridge (host tools, called by the agent running inside the sandbox). The system prompt forces both before any repo work, so a broken bridge is obvious:

  • fetchIssueComments — a normal server tool (host fetches the issue thread).
  • searchRelatedIssues — exposed via code mode: the agent calls execute_typescript, and the generated code invokes external_searchRelatedIssues(...) (run on a host isolate).

The bridge is a localhost HTTP server, so it's only directly reachable from same-machine providers (local process, Docker). For remote cloud sandboxes (Daytona, Vercel) the agent can't dial your laptop — set NGROK_AUTHTOKEN and the example tunnels the bridge out over ngrok so the tools work there too (the per-run bearer token still gates every call). Without it, cloud runs skip the tools and do a plain triage. In production you wouldn't need ngrok — your orchestrator already has a public URL to advertise.

Set keys in .env.local, then pnpm dev and open /sandboxes.