configure-chatkit
About
This skill guides developers through integrating OpenAI ChatKit into Next.js applications, including environment configuration, UI component building, and API endpoint integration. It's used when setting up chatbot frontends that display conversation history and tool call results visually. Key capabilities include ChatKit UI implementation, domain allowlist management, and Better Auth session integration.
Quick Install
Claude Code
Recommended/plugin add https://github.com/majiayu000/claude-skill-registrygit clone https://github.com/majiayu000/claude-skill-registry.git ~/.claude/skills/configure-chatkitCopy and paste this command in Claude Code to install this skill
Documentation
Configure ChatKit Skill
This skill provides guidance for integrating OpenAI ChatKit in a Next.js frontend.
Purpose
Setup OpenAI ChatKit in Next.js:
- Add domain allowlist key to environment
- Build chat UI component
- POST to /api/chat endpoint with conversation_id
- Display conversation history and tool calls visually
- Integrate Better Auth session
When to Use
Use this skill when:
- Building the chatbot frontend UI
- Integrating ChatKit components
- Setting up chat API communication
- Displaying tool execution results
- Connecting frontend with authenticated backend
Capabilities
- ChatKit Integration: Use OpenAI ChatKit for chat UI
- Environment Configuration: Set up domain allowlist
- API Communication: POST to chat endpoint with conversation context
- Visual Feedback: Display history, tool calls, and results
- Authentication: Integrate Better Auth session
Environment Setup
# .env.local
NEXT_PUBLIC_CHAT_KIT_PUBLIC_KEY=your_chatkit_public_key
CHAT_KIT_SECRET=your_chatkit_secret_key
CHAT_KIT_INSTANCE_ID=your_instance_id
NEXT_PUBLIC_ALLOWED_DOMAINS=localhost:3000,yourdomain.com
Implementation Pattern
Chat Component
"use client";
import { useState, useEffect, useRef } from "react";
import { useChat } from "@openai/chatkit";
import { useSession } from "better-auth/react";
interface ChatWidgetProps {
conversationId?: string;
}
export function ChatWidget({ conversationId }: ChatWidgetProps) {
const { data: session } = useSession();
const [input, setInput] = useState("");
const [isLoading, setIsLoading] = useState(false);
const messagesEndRef = useRef<HTMLDivElement>(null);
const {
messages,
sendMessage,
isLoading: chatLoading,
error
} = useChat({
apiUrl: "/api/chat",
conversationId,
auth: {
getToken: () => session?.accessToken
}
});
const handleSend = async () => {
if (!input.trim() || isLoading) return;
setIsLoading(true);
try {
await sendMessage(input);
setInput("");
} catch (err) {
console.error("Failed to send message:", err);
} finally {
setIsLoading(false);
}
};
return (
<div className="chat-container">
<div className="chat-messages">
{messages.map((msg) => (
<MessageBubble
key={msg.id}
role={msg.role}
content={msg.content}
toolCalls={msg.tool_calls}
/>
))}
<div ref={messagesEndRef} />
</div>
<div className="chat-input">
<input
value={input}
onChange={(e) => setInput(e.target.value)}
onKeyDown={(e) => e.key === "Enter" && handleSend()}
placeholder="Ask me to manage your tasks..."
disabled={isLoading}
/>
<button onClick={handleSend} disabled={isLoading}>
{isLoading ? "Sending..." : "Send"}
</button>
</div>
{error && <ErrorToast message={error.message} />}
</div>
);
}
Message Bubble with Tool Calls
interface MessageBubbleProps {
role: "user" | "assistant" | "tool";
content: string;
toolCalls?: ToolCall[];
}
function MessageBubble({ role, content, toolCalls }: MessageBubbleProps) {
const isUser = role === "user";
return (
<div className={`message ${isUser ? "user" : "assistant"}`}>
<div className="message-content">{content}</div>
{toolCalls && toolCalls.length > 0 && (
<div className="tool-calls">
{toolCalls.map((call) => (
<ToolCallBadge key={call.id} tool={call} />
))}
</div>
)}
</div>
);
}
function ToolCallBadge({ tool }: { tool: ToolCall }) {
return (
<div className="tool-badge">
<span className="tool-icon">🔧</span>
<span className="tool-name">{tool.name}</span>
{tool.status === "completed" && <span className="check">✓</span>}
{tool.status === "failed" && <span className="x">✗</span>}
</div>
);
}
API Communication
async function sendMessage(message: string): Promise<void> {
const response = await fetch("/api/chat", {
method: "POST",
headers: {
"Content-Type": "application/json",
"Authorization": `Bearer ${session?.accessToken}`
},
body: JSON.stringify({
message,
conversation_id: conversationId
})
});
if (!response.ok) {
throw new Error("Failed to send message");
}
const data = await response.json();
// Update conversation_id if new
if (!conversationId && data.conversation_id) {
setConversationId(data.conversation_id);
}
}
Better Auth Integration
"use client";
import { SessionProvider } from "better-auth/react";
export function AuthProvider({ children }: { children: React.ReactNode }) {
return (
<SessionProvider>
{children}
</SessionProvider>
);
}
// In your chat component
const { data: session } = useSession();
useEffect(() => {
if (session?.accessToken) {
// Configure chat client with auth token
chatClient.setAuthToken(session.accessToken);
}
}, [session]);
Styling (Tailwind CSS)
.chat-container {
@apply flex flex-col h-full max-w-2xl mx-auto border rounded-lg shadow-lg;
}
.chat-messages {
@apply flex-1 overflow-y-auto p-4 space-y-4;
}
.message {
@apply p-3 rounded-lg max-w-[80%];
}
.message.user {
@apply bg-blue-500 text-white ml-auto;
}
.message.assistant {
@apply bg-gray-100 text-gray-900;
}
.chat-input {
@apply border-t p-4 flex gap-2;
}
.chat-input input {
@apply flex-1 px-4 py-2 border rounded-lg focus:outline-none focus:ring-2;
}
.chat-input button {
@apply px-4 py-2 bg-blue-500 text-white rounded-lg hover:bg-blue-600;
}
.tool-calls {
@apply mt-2 flex flex-wrap gap-2;
}
.tool-badge {
@apply inline-flex items-center gap-1 px-2 py-1 bg-yellow-100 text-yellow-800 text-xs rounded;
}
.error-toast {
@apply absolute bottom-20 right-4 bg-red-500 text-white px-4 py-2 rounded-lg shadow;
}
Error Handling
function ErrorToast({ message }: { message: string }) {
return (
<div className="error-toast">
<span>Error: {message}</span>
<button onClick={() => clearError()}>Dismiss</button>
</div>
);
}
function LoadingIndicator() {
return (
<div className="typing-indicator">
<span></span>
<span></span>
<span></span>
</div>
);
}
Verification Checklist
- ChatKit configured with allowed domains
- Chat widget displays messages correctly
- Messages sent to /api/chat endpoint
- Conversation ID maintained across messages
- Tool calls displayed visually
- Better Auth session integrated
- Error toasts shown on failures
- Loading indicators during API calls
GitHub Repository
Related Skills
content-collections
MetaThis skill provides a production-tested setup for Content Collections, a TypeScript-first tool that transforms Markdown/MDX files into type-safe data collections with Zod validation. Use it when building blogs, documentation sites, or content-heavy Vite + React applications to ensure type safety and automatic content validation. It covers everything from Vite plugin configuration and MDX compilation to deployment optimization and schema validation.
creating-opencode-plugins
MetaThis skill provides the structure and API specifications for creating OpenCode plugins that hook into 25+ event types like commands, files, and LSP operations. It offers implementation patterns for JavaScript/TypeScript modules that intercept and extend the AI assistant's lifecycle. Use it when you need to build event-driven plugins for monitoring, custom handling, or extending OpenCode's capabilities.
sglang
MetaSGLang is a high-performance LLM serving framework that specializes in fast, structured generation for JSON, regex, and agentic workflows using its RadixAttention prefix caching. It delivers significantly faster inference, especially for tasks with repeated prefixes, making it ideal for complex, structured outputs and multi-turn conversations. Choose SGLang over alternatives like vLLM when you need constrained decoding or are building applications with extensive prefix sharing.
evaluating-llms-harness
TestingThis Claude Skill runs the lm-evaluation-harness to benchmark LLMs across 60+ standardized academic tasks like MMLU and GSM8K. It's designed for developers to compare model quality, track training progress, or report academic results. The tool supports various backends including HuggingFace and vLLM models.
