TanStack AI

https://tanstack.com/ai/latest
Admin
TanStack AI is a powerful, open-source AI SDK providing a unified, type-safe interface across...
Tokens:475,380
Snippets:4,834
Trust Score:10
Update:1 month ago
Show doc for...
Tokens:
Raw
### Run the Rerank Example Application

Source: https://tanstack.com/ai/latest/docs/rerank/rerank.md

Commands to install dependencies and start the local development server for the rerank example app.

```bash
cd examples/ts-react-rerank
pnpm install
cp .env.example .env   # add COHERE_API_KEY and/or OPENROUTER_API_KEY
pnpm dev
```

--------------------------------

### Run the Expo Recipe Example

Source: https://tanstack.com/ai/latest/docs/getting-started/quick-start-react-native.md

Execute the development command to start the Hono server and Expo/Metro environment.

```bash
pnpm --filter ts-react-native-chat dev
```

--------------------------------

### Run Development Server

Source: https://tanstack.com/ai/latest/docs/getting-started/quick-start-react-native.md

Install tsx and start the Hono development server.

```bash
pnpm add -D tsx
pnpm pkg set scripts.dev:server="tsx server.ts"
pnpm dev:server
```

--------------------------------

### Install Persistence Package

Source: https://tanstack.com/ai/latest/docs/persistence/build-your-own-adapter.md

Install the necessary persistence package and run the intent installer to set up your environment.

```bash
pnpm add @tanstack/ai-persistence
npx @tanstack/intent@latest install
```

--------------------------------

### Install Groq Adapter

Source: https://tanstack.com/ai/latest/docs/adapters/groq.md

Install the required package via npm.

```bash
npm install @tanstack/ai-groq
```

--------------------------------

### Install @tanstack/ai-client

Source: https://tanstack.com/ai/latest/docs/api/ai-client.md

Use npm to install the package in your project.

```bash
npm install @tanstack/ai-client
```

--------------------------------

### Install OpenAI Adapter

Source: https://tanstack.com/ai/latest/docs/adapters/openai-compatible.md

Install the package containing the compatible adapter.

```bash
npm install @tanstack/ai-openai
```

--------------------------------

### Install Vertex AI Adapter

Source: https://tanstack.com/ai/latest/docs/adapters/vertex.md

Command to install the required package.

```bash
npm install @tanstack/ai-vertex
```

--------------------------------

### Usage Example for resumeHttpResponse

Source: https://tanstack.com/ai/latest/docs/reference/functions/resumeHttpResponse

Example implementation of a GET request handler using resumeHttpResponse.

```typescript
export async function GET(request: Request) {
  return resumeHttpResponse({ adapter: memoryStream(request) });
}
```

--------------------------------

### Install OpenCode adapter

Source: https://tanstack.com/ai/latest/docs/adapters/opencode.md

Install the required package for the OpenCode adapter.

```bash
npm install @tanstack/ai-opencode
```

--------------------------------

### Install Gemini Adapter

Source: https://tanstack.com/ai/latest/docs/adapters/gemini.md

Install the required package via npm.

```bash
npm install @tanstack/ai-gemini
```

--------------------------------

### Install Dependencies

Source: https://tanstack.com/ai/latest/docs/getting-started/quick-start-react-native.md

Install the necessary TanStack AI packages, Hono, and OpenAI adapter.

```bash
pnpm add @tanstack/ai @tanstack/ai-react @tanstack/ai-openai hono @hono/node-server zod
```

--------------------------------

### Install OpenRouter Adapter

Source: https://tanstack.com/ai/latest/docs/adapters/openrouter.md

Install the package via npm.

```bash
npm install @tanstack/ai-openrouter
```

--------------------------------

### Install @tanstack/ai-preact

Source: https://tanstack.com/ai/latest/docs/api/ai-preact.md

Install the package using npm.

```bash
npm install @tanstack/ai-preact
```

--------------------------------

### Install Cloudflare Adapter and Dependencies

Source: https://tanstack.com/ai/latest/docs/community-adapters/cloudflare.md

Install the core adapter and optional provider SDKs for AI Gateway.

```bash
npm install @cloudflare/tanstack-ai @tanstack/ai
```

```bash
npm install @tanstack/ai-openai       # For OpenAI via Gateway
npm install @tanstack/ai-anthropic    # For Anthropic via Gateway
npm install @tanstack/ai-gemini       # For Gemini via Gateway
npm install @tanstack/ai-grok         # For Grok via Gateway
npm install @tanstack/ai-openrouter   # For OpenRouter via Gateway
```

--------------------------------

### Install Bedrock Adapter

Source: https://tanstack.com/ai/latest/docs/adapters/bedrock.md

Install the required package for the Bedrock adapter.

```bash
pnpm add @tanstack/ai-bedrock
```

--------------------------------

### Install Cohere Adapter

Source: https://tanstack.com/ai/latest/docs/adapters/cohere.md

Install the necessary packages for using the Cohere adapter.

```bash
npm install @tanstack/ai @tanstack/ai-cohere
```

--------------------------------

### Install and authenticate OpenCode CLI

Source: https://tanstack.com/ai/latest/docs/adapters/opencode.md

Install the OpenCode CLI globally and authenticate the host.

```bash
npm install -g opencode-ai
opencode auth login
```

--------------------------------

### Install Soniox Adapter

Source: https://tanstack.com/ai/latest/docs/community-adapters/soniox.md

Install the required package via npm.

```bash
npm install @soniox/tanstack-ai-adapter
```

--------------------------------

### Install @tanstack/ai-solid

Source: https://tanstack.com/ai/latest/docs/api/ai-solid.md

Install the package via npm.

```bash
npm install @tanstack/ai-solid
```

--------------------------------

### Install @tanstack/ai-svelte

Source: https://tanstack.com/ai/latest/docs/api/ai-svelte.md

Install the package using npm.

```bash
npm install @tanstack/ai-svelte
```

--------------------------------

### Install the memory package

Source: https://tanstack.com/ai/latest/docs/memory/quickstart.md

Install the required dependency using pnpm.

```bash
pnpm add @tanstack/ai-memory
```

--------------------------------

### Install ElevenLabs Adapter

Source: https://tanstack.com/ai/latest/docs/adapters/elevenlabs.md

Install the adapter package and its required peer dependencies.

```bash
npm install @tanstack/ai-elevenlabs
```

```bash
npm install @tanstack/ai @tanstack/ai-client
```

--------------------------------

### Install QuickJS Bun Driver

Source: https://tanstack.com/ai/latest/docs/code-mode/code-mode-isolates.md

Install the package using the Bun package manager.

```bash
bun add @tanstack/ai-isolate-quickjs-bun
```

--------------------------------

### Install Cencori SDK

Source: https://tanstack.com/ai/latest/docs/community-adapters/cencori.md

Install the required package via npm.

```bash
npm install @cencori/ai-sdk
```

--------------------------------

### Install MCP dependencies

Source: https://tanstack.com/ai/latest/docs/tools/mcp.md

Install the required MCP client and SDK packages.

```bash
pnpm add @tanstack/ai-mcp @modelcontextprotocol/sdk
```

--------------------------------

### Install Ollama Adapter

Source: https://tanstack.com/ai/latest/docs/adapters/ollama.md

Install the required package via npm.

```bash
npm install @tanstack/ai-ollama
```

--------------------------------

### Install Grok adapter

Source: https://tanstack.com/ai/latest/docs/adapters/grok.md

Install the required package for Grok integration.

```bash
npm install @tanstack/ai-grok
```

--------------------------------

### Install Reranking Packages

Source: https://tanstack.com/ai/latest/docs/rerank/rerank.md

Install the necessary adapter and core packages via npm.

```bash
npm install @tanstack/ai-cohere
# or, to rerank through OpenRouter:
npm install @tanstack/ai-openrouter
```

```bash
npm install @tanstack/ai
```

--------------------------------

### Install the fal.ai adapter

Source: https://tanstack.com/ai/latest/docs/adapters/fal.md

Use npm to install the required package for the fal.ai adapter.

```bash
npm install @tanstack/ai-fal
```

--------------------------------

### Install Persistence Package

Source: https://tanstack.com/ai/latest/docs/persistence/overview.md

Install the required persistence package using pnpm.

```bash
pnpm add @tanstack/ai-persistence
```

--------------------------------

### Install Mistral Adapter

Source: https://tanstack.com/ai/latest/docs/adapters/mistral.md

Install the required package for Mistral integration.

```bash
npm install @tanstack/ai-mistral
```

--------------------------------

### Install LLM Gateway Adapter

Source: https://tanstack.com/ai/latest/docs/adapters/llmgateway.md

Install the required package via npm.

```bash
npm install @tanstack/ai-llmgateway
```

--------------------------------

### Install TanStack AI packages

Source: https://tanstack.com/ai/latest/docs/sandbox/quick-start.md

Install the core library, Grok Build adapter, sandbox utilities, and Docker provider.

```bash
npm i @tanstack/ai @tanstack/ai-grok-build @tanstack/ai-sandbox @tanstack/ai-sandbox-docker
```

--------------------------------

### Install TanStack AI Devtools

Source: https://tanstack.com/ai/latest/docs/getting-started/devtools.md

Install the required devtools packages for React, SolidJS, or Preact.

```bash
npm install -D @tanstack/react-ai-devtools @tanstack/react-devtools
```

```bash
npm install -D @tanstack/solid-ai-devtools @tanstack/solid-devtools
```

```bash
npm install -D @tanstack/preact-ai-devtools @tanstack/preact-devtools
```

--------------------------------

### Install local process sandbox package

Source: https://tanstack.com/ai/latest/docs/sandbox/quick-start.md

Install the required package to run the agent directly on the host machine.

```bash
npm i @tanstack/ai-sandbox-local-process
```

--------------------------------

### Install BytePlus Adapter

Source: https://tanstack.com/ai/latest/docs/adapters/byteplus.md

Commands to install the @tanstack/ai-byteplus package using common package managers.

```bash
npm install @tanstack/ai-byteplus
# or
pnpm add @tanstack/ai-byteplus
# or
yarn add @tanstack/ai-byteplus
```

--------------------------------

### Install TanStack AI packages

Source: https://tanstack.com/ai/latest/docs/code-mode/code-mode.md

Install the core packages and select an appropriate isolate driver based on your runtime environment.

```bash
pnpm add @tanstack/ai @tanstack/ai-code-mode zod
```

```bash
# Node.js — fastest, uses V8 isolates (requires native compilation)
pnpm add @tanstack/ai-isolate-node

# QuickJS WASM — no native deps, works in browsers and edge runtimes
pnpm add @tanstack/ai-isolate-quickjs

# QuickJS Bun — native QuickJS via bun:ffi, fastest option on Bun
bun add @tanstack/ai-isolate-quickjs-bun

# Cloudflare Workers — run on the edge
pnpm add @tanstack/ai-isolate-cloudflare

# Daytona sandboxes — run in a remote Daytona sandbox
pnpm add @tanstack/ai-isolate-daytona @daytona/sdk
```

--------------------------------

### Install TanStack AI dependencies

Source: https://tanstack.com/ai/latest/docs/getting-started/quick-start-server.md

Install the core library and the OpenAI adapter using your preferred package manager.

```bash
npm install @tanstack/ai @tanstack/ai-openai
# or
pnpm add @tanstack/ai @tanstack/ai-openai
# or
yarn add @tanstack/ai @tanstack/ai-openai
```

--------------------------------

### Install ACP Adapter and Sandbox Dependencies

Source: https://tanstack.com/ai/latest/docs/adapters/acp-compatible.md

Install the required packages for ACP compatibility and sandbox execution.

```bash
npm install @tanstack/ai-acp @tanstack/ai @tanstack/ai-sandbox @tanstack/ai-sandbox-docker
```

--------------------------------

### Usage of resumeServerSentEventsResponse in a GET handler

Source: https://tanstack.com/ai/latest/docs/reference/functions/resumeServerSentEventsResponse

Example of implementing a GET handler that resumes a server-sent events response using a memory stream adapter.

```typescript
export async function GET(request: Request) {
  return resumeServerSentEventsResponse({ adapter: memoryStream(request) });
}
```

--------------------------------

### setup(ctx)

Source: https://tanstack.com/ai/latest/docs/reference/interfaces/ChatMiddleware

The setup method is a provisioning hook that runs before onConfig. It is used to call provide accessors to make capabilities available to subsequent middleware.

```APIDOC
## setup(ctx)

### Description
Provisioning hook that runs first, before onConfig, across all middleware in array order. Use it to call provide accessors so later middleware can consume the capabilities.

### Parameters
- **ctx** (ChatMiddlewareContext<TContext>) - Required - The stable context object.

### Returns
- **void | Promise<void>** - Returns nothing or a promise that resolves when setup is complete.
```

--------------------------------

### setup

Source: https://tanstack.com/ai/latest/docs/reference/interfaces/DefinedChatMiddleware

A provisioning hook that runs before onConfig to allow middleware to provide accessors for subsequent use.

```APIDOC
## setup

### Description
Provisioning hook. Runs first, before `onConfig` (init), across all middleware in array order. Use it to call `provide` accessors so later middleware can consume the capabilities. Receives the stable context but does not receive the mutable config.

### Signature
`optional setup?: (ctx: ChatMiddlewareContext<TContext>) => void | Promise<void>;`

### Parameters
- **ctx** (ChatMiddlewareContext<TContext>) - Required - The stable chat middleware context.
```

--------------------------------

### Initialize Expo Project

Source: https://tanstack.com/ai/latest/docs/getting-started/quick-start-react-native.md

Create a new Expo application from scratch.

```bash
npx create-expo-app@latest my-ai-chat
```

--------------------------------

### Implementing BlobStore Range Requests

Source: https://tanstack.com/ai/latest/docs/persistence/store-reference.md

Example implementation of the get method using resolveBlobRange to handle partial content requests.

```ts
import { resolveBlobRange } from '@tanstack/ai-persistence'

async get(key: string, options?: BlobGetOptions) {
  const row = await selectBlob(key)
  if (!row) return null
  if (!options?.range) return blobObject(row, row.body)
  const served = resolveBlobRange(row.size, options.range)
  // Slice at the storage layer, not after loading the whole object.
  const bytes = await selectBlobSlice(key, served.offset, served.length)
  return blobObject(row, bytes, served)
}
```

--------------------------------

### Usage in Frameworks with Native Response Handling

Source: https://tanstack.com/ai/latest/docs/reference/functions/chatParamsFromRequest

Example implementation for frameworks that automatically handle thrown Response objects, such as TanStack Start or Remix.

```ts
export async function POST(req: Request) {
  const params = await chatParamsFromRequest(req)
  // ...use params
}
```

--------------------------------

### Start Ollama server

Source: https://tanstack.com/ai/latest/docs/adapters/ollama.md

Launch the local Ollama server instance.

```bash
ollama serve
```

--------------------------------

### Install @tanstack/ai-vue

Source: https://tanstack.com/ai/latest/docs/api/ai-vue.md

Install the package using npm.

```bash
npm install @tanstack/ai-vue
```

--------------------------------

### Initialize TanStack Intent

Source: https://tanstack.com/ai/latest/docs/getting-started/agent-skills.md

Run the intent installation command from the project root.

```bash
npx @tanstack/intent@latest install
```

--------------------------------

### Configure ACP-Compatible Harness

Source: https://tanstack.com/ai/latest/docs/adapters/acp-compatible.md

Examples showing how to initialize the adapter using either API key authentication or host-based CLI login.

```ts
import { acpCompatibleText } from "@tanstack/ai-acp"

acpCompatibleText("composer-2.5", {
  name: "acp",
  command: ({ model }) => `grok agent -m '${model}' --always-approve stdio`,
  authMethodId: "xai.api_key",
})
```

```ts
acpCompatibleText("composer-2.5", {
  name: "acp",
  command: ({ model }) => `grok agent -m '${model}' --always-approve stdio`,
  authMode: "host",
})
```

--------------------------------

### Install @tanstack/ai-react

Source: https://tanstack.com/ai/latest/docs/api/ai-react.md

Command to install the package via npm.

```bash
npm install @tanstack/ai-react
```

--------------------------------

### onStart

Source: https://tanstack.com/ai/latest/docs/advanced/middleware.md

A middleware hook called once after the initial configuration phase completes, suitable for setup tasks.

```APIDOC
## onStart

### Description
Called once after the initial `onConfig` completes. Use it for setup tasks like initializing timers or logging.

### Parameters
- **ctx** (Context) - The current request context, including `requestId` and `iteration`.

### Example
```typescript
import { type ChatMiddleware } from "@tanstack/ai";

const timer: ChatMiddleware = {
  name: "timer",
  onStart: (ctx) => {
    console.log(`Request ${ctx.requestId} started at iteration ${ctx.iteration}`);
  },
};
```
```

--------------------------------

### Basic Usage Example

Source: https://tanstack.com/ai/latest/docs/reference/functions/resumeWebSocketResponse

A simple implementation of resumeWebSocketResponse using a memory stream adapter.

```ts
resumeWebSocketResponse({ adapter: memoryStream(request) })
```

--------------------------------

### Configure environment variables

Source: https://tanstack.com/ai/latest/docs/getting-started/quick-start-server.md

Set up your API keys in a .env file for the adapter to read at runtime.

```bash
# OpenRouter (recommended — access 300+ models with one key)
OPENROUTER_API_KEY=sk-or-...

# OpenAI
OPENAI_API_KEY=your-openai-api-key
```

--------------------------------

### Install Decart Adapter

Source: https://tanstack.com/ai/latest/docs/community-adapters/decart.md

Install the required package via npm.

```bash
npm install @decartai/tanstack-ai-adapter
```

--------------------------------

### Install @tanstack/ai-angular

Source: https://tanstack.com/ai/latest/docs/api/ai-angular.md

Use npm to install the package in your Angular project.

```bash
npm install @tanstack/ai-angular
```

--------------------------------

### startAudioCapture()

Source: https://tanstack.com/ai/latest/docs/reference/interfaces/RealtimeConnection.md

Starts capturing audio from the microphone.

```APIDOC
## startAudioCapture()

### Description
Starts capturing audio from the microphone.

### Returns
- **Promise<void>** - A promise that resolves when audio capture has started.
```

--------------------------------

### Install @tanstack/ai-octane

Source: https://tanstack.com/ai/latest/docs/api/ai-octane.md

Install the package along with its required peer dependency.

```bash
npm install @tanstack/ai-octane octane
```

--------------------------------

### Use SystemPrompt in chat configuration

Source: https://tanstack.com/ai/latest/docs/reference/type-aliases/SystemPrompt

Examples demonstrating both simple string usage and provider-specific metadata configuration.

```ts
// The 90% case — plain strings work everywhere.
  systemPrompts: ['Be concise.', 'Cite sources.']
```

```ts
// Provider-specific metadata via the object form. No `satisfies` cast
  // is needed — the adapter narrows the `metadata` field's type at the
  // call site so users get autocomplete and structural checking
  // automatically.
  import { anthropicText } from '@tanstack/ai-anthropic'

  chat({
    adapter: anthropicText(),
    systemPrompts: [
      {
        content: 'Stable instructions — cache me.',
        metadata: { cache_control: { type: 'ephemeral' } },
      },
      'Volatile per-request instruction.',
    ],
  })
```

--------------------------------

### Initialize Adapter with Explicit API Key

Source: https://tanstack.com/ai/latest/docs/media/video-generation.md

Shows how to instantiate a video adapter by passing an API key directly, bypassing environment variable lookups.

```typescript
import { createOpenaiVideo } from "@tanstack/ai-openai";

const adapter = createOpenaiVideo("sora-2", "your-openai-api-key");
```

--------------------------------

### Install Ollama

Source: https://tanstack.com/ai/latest/docs/adapters/ollama.md

Commands to install Ollama on macOS, Linux, and Windows.

```bash
# macOS
brew install ollama

# Linux
curl -fsSL https://ollama.com/install.sh | sh

# Windows
# Download from https://ollama.com
```

--------------------------------

### Install Codex Adapter

Source: https://tanstack.com/ai/latest/docs/adapters/codex.md

Install the required package via npm.

```bash
npm install @tanstack/ai-codex
```

--------------------------------

### Install Anthropic Adapter

Source: https://tanstack.com/ai/latest/docs/adapters/anthropic.md

Install the required package via npm.

```bash
npm install @tanstack/ai-anthropic
```

--------------------------------

### Configure MCP client with tool approval logic

Source: https://tanstack.com/ai/latest/docs/tools/mcp.md

Demonstrates initializing an MCP client and using tool metadata annotations to conditionally set approval requirements based on server trust.

```ts
import { createMCPClient } from '@tanstack/ai-mcp'

const url = 'https://my-mcp-server.example.com/mcp'

// Trust comes from YOUR configuration — an allowlist of servers you operate or
// have vetted — never from anything the server itself sends.
const trustedServers = new Set(['https://my-mcp-server.example.com/mcp'])
const serverIsTrusted = trustedServers.has(url)

const mcp = await createMCPClient({ transport: { type: 'http', url } })

const tools = (await mcp.tools()).map((tool) => {
  const meta = tool.metadata.mcp
  const advertisedReadOnly = meta.annotations?.readOnlyHint === true
  return {
    ...tool,
    // Approval is the default. A hint may only relax it for a server whose
    // trust you established independently; on any other server the same hint
    // is a label/recommendation and changes nothing about approval.
    needsApproval: !(serverIsTrusted && advertisedReadOnly),
  }
})
```

--------------------------------

### Initialize QuickJS Driver

Source: https://tanstack.com/ai/latest/docs/code-mode/code-mode-isolates.md

Basic configuration for the QuickJS driver, including memory, timeout, and stack size limits.

```typescript
import { createQuickJSIsolateDriver } from '@tanstack/ai-isolate-quickjs'

const driver = createQuickJSIsolateDriver({
  memoryLimit: 128,     // MB
  timeout: 30_000,      // ms
  maxStackSize: 524288, // bytes (512 KiB)
  wasmLocation: '/assets/quickjs/emscripten-module.wasm',
})
```

--------------------------------

### mem0()

Source: https://tanstack.com/ai/latest/docs/memory/adapters.md

Initializes a hosted adapter backed by a mem0 server.

```APIDOC
## mem0()

### Description
Initializes a hosted adapter backed by a mem0 server. This adapter requires a running mem0 server and communicates over HTTP.

### Parameters
- **user** (string) - Optional - mem0 user_id. Defaults to scope.userId or 'demo-user'.
- **baseUrl** (string) - Optional - Server URL. Defaults to MEM0_URL or 'http://localhost:8000'.
- **apiKey** (string) - Optional - Bearer token. Defaults to MEM0_ADMIN_API_KEY.
- **rerank** (boolean) - Optional - Ask mem0 to rerank search results. Defaults to true.
- **threshold** (number) - Optional - Minimum search score. Defaults to 0.1.

### Usage Example
```ts
import { mem0 } from '@tanstack/ai-memory/mem0'

const memory = mem0({
  user: 'alice',
  baseUrl: 'https://mem0.internal',
  apiKey: process.env.MEM0_ADMIN_API_KEY,
  rerank: true,
  threshold: 0.2,
})
```
```

--------------------------------

### Install OpenTelemetry dependency

Source: https://tanstack.com/ai/latest/docs/advanced/otel.md

Install the required peer dependency for OpenTelemetry support.

```bash
pnpm add @opentelemetry/api
```

--------------------------------

### Implementing Chat with OpenAI Adapter

Source: https://tanstack.com/ai/latest/docs/advanced/tree-shaking.md

Example of using only the chat activity and the OpenAI text adapter to minimize bundle size.

```ts
// Only chat code is bundled
import { chat } from '@tanstack/ai'
import { openaiText } from '@tanstack/ai-openai'

const stream = chat({
  adapter: openaiText('gpt-5.5'),
  messages: [{ role: 'user', content: 'Hello!' }],
})
```

--------------------------------

### bedrockText(model, options?)

Source: https://tanstack.com/ai/latest/docs/adapters/bedrock.md

Initializes a Bedrock adapter using the Converse API, which is the recommended path for most models including Claude, Nova, and Llama.

```APIDOC
## bedrockText(model, options?)

### Description
Creates a `bedrock-converse` adapter backed by `@aws-sdk/client-bedrock-runtime`. This is the recommended API for most use cases.

### Parameters
- **model** (string) - Required - The model ID (e.g., 'us.anthropic.claude-haiku-4-5-20251001-v1:0').
- **options** (object) - Optional - Configuration object.
  - **region** (string) - Optional - AWS region.
  - **api** (string) - Optional - Set to 'converse' (default) or 'chat'.
```

--------------------------------

### Usage example for createChatOptions

Source: https://tanstack.com/ai/latest/docs/reference/functions/createChatOptions

Demonstrates how to pre-define chat options and pass them into the chat function.

```ts
const chatOptions = createChatOptions({
  adapter: anthropicText('claude-sonnet-4-5'),
})

const stream = chat({ ...chatOptions, messages })
```

--------------------------------

### Install Claude Code Adapter

Source: https://tanstack.com/ai/latest/docs/adapters/claude-code.md

Install the required package via npm.

```bash
npm install @tanstack/ai-claude-code
```

--------------------------------

### Initialize QuickJS Bun Driver

Source: https://tanstack.com/ai/latest/docs/code-mode/code-mode-isolates.md

Create a driver instance with custom memory, timeout, and stack size constraints.

```typescript
import { createQuickJSBunIsolateDriver } from '@tanstack/ai-isolate-quickjs-bun'

const driver = createQuickJSBunIsolateDriver({
  memoryLimit: 128,     // MB
  timeout: 30_000,      // ms
  maxStackSize: 524288, // bytes (512 KiB)
})
```

--------------------------------

### Install MCP App dependencies

Source: https://tanstack.com/ai/latest/docs/mcp/apps.md

Install the required packages for MCP app integration.

```bash
pnpm add @tanstack/ai-mcp @tanstack/ai-client @mcp-ui/client
```

--------------------------------

### onStart

Source: https://tanstack.com/ai/latest/docs/reference/interfaces/ChatMiddleware

Called when the chat run starts, immediately following the initial onConfig call.

```APIDOC
## onStart(ctx)

### Description
Called when the chat run starts, after the initial onConfig execution.

### Parameters
- **ctx** (ChatMiddlewareContext) - The current chat middleware context.

### Returns
- (void | Promise<void>)
```
Tokens:
Raw