Clarity Engine SDK

One message → clean, structured, actionable plans.

Proactive AI clarity coach that auto-relabels messy input, builds a live hierarchical tree, and leads users to full clarity. White-label ready for AI companies.

Install

npm install clarity-engine-sdk openai
# Optional: Supabase for vector memory
npm install @supabase/supabase-js
# Optional: React components
npm install react react-dom

Quick Start

Node.js / Server

import { ClarityEngine } from "clarity-engine-sdk";

const engine = new ClarityEngine({
  openaiApiKey: process.env.OPENAI_API_KEY!,
  model: "gpt-4o",
  onClarityUpdate: (score, tree) => {
    console.log(`Clarity: ${score}/100`);
  },
});

const session = await engine.startSession();

const result = await engine.processMessage(
  session.id,
  "I need to launch my app but I have no idea where to start, everything feels overwhelming"
);

console.log(result.response);
console.log(`Clarity score: ${result.session.clarityScore}/100`);
console.log(engine.getTreeMarkdown(session.id));

React Hook

import { useClarity } from "clarity-engine-sdk";

function MyApp() {
  const { messages, sendMessage, clarityScore, isLoading, startSession } = useClarity({
    openaiApiKey: process.env.OPENAI_API_KEY!,
  });

  useEffect(() => { startSession(); }, []);

  return (
    <div>
      <div>Clarity Score: {clarityScore}/100</div>
      {messages.map(msg => (
        <div key={msg.id} style={{ textAlign: msg.role === "user" ? "right" : "left" }}>
          {msg.content}
        </div>
      ))}
      <button onClick={() => sendMessage("I want to build a startup")} disabled={isLoading}>
        Send
      </button>
    </div>
  );
}

Drop-in Chat Component

import { ClarityChat } from "clarity-engine-sdk";

function App() {
  return (
    <div style={{ height: 600 }}>
      <ClarityChat
        config={{ openaiApiKey: process.env.OPENAI_API_KEY! }}
        welcomeMessage="What's on your mind? Let's turn it into a plan."
        showClarityScore
        onComplete={(markdown) => console.log("Plan ready:\n", markdown)}
      />
    </div>
  );
}

White-Label Branding

const engine = new ClarityEngine({
  openaiApiKey: "...",
  branding: {
    name: "Aria",
    persona: "You speak in a warm, coach-like tone.",
  },
});

<ClarityChat
  config={{
    openaiApiKey: "...",
    branding: { name: "Aria" },
  }}
  theme={{
    primary: "#f97316",
    background: "#0a0a0a",
    surface: "#161616",
  }}
/>

API Reference

ClarityEngine

| Method | Description | |--------|-------------| | startSession(options?) | Start a new clarity session | | processMessage(sessionId, message) | Send a message and receive a structured response | | getSession(sessionId) | Retrieve the current session state | | getTreeMarkdown(sessionId) | Export the clarity tree as markdown | | clearSession(sessionId) | End and remove a session |

ClarityEngineConfig

| Property | Type | Description | |----------|------|-------------| | openaiApiKey | string | Required. OpenAI API key | | model | string | Model to use (default: gpt-4o) | | temperature | number | Sampling temperature (default: 0.7) | | maxTokens | number | Max tokens per response (default: 1024) | | supabaseUrl | string | Optional Supabase URL for vector memory | | supabaseKey | string | Optional Supabase anon key | | branding | object | White-label name and persona | | onClarityUpdate | function | Callback on clarity score changes | | onSessionComplete | function | Callback when session reaches 100 clarity |

useClarity(config)

Returns: { messages, session, tree, clarityScore, phase, isLoading, isComplete, error, startSession, sendMessage, reset, exportMarkdown }

License

Commercial license — see LICENSE.

Personal / internal use: free.
White-label / resell in your product: $4,999 one-time.
Contact: [email protected]