CLS++ JavaScript/TypeScript SDK

Brain-inspired, model-agnostic persistent memory for Large Language Models.

Installation

npm install clsplusplus

Get an API key

Every call needs an API key — without one the first learn() returns 401.

1. Sign up at https://www.clsplusplus.com and open your profile page.
2. Copy your key (cls_live_…) and expose it to the SDK:

export CLS_API_KEY="cls_live_xxxxxxxxxxxx"

Or pass it explicitly — see Configuration below.

Quick Start

import { Brain } from "clsplusplus";

const brain = new Brain("alice");

// Teach it anything
await brain.learn("I work at Google as a senior engineer");
await brain.learn("I prefer Python over JavaScript");

// Ask it anything — semantic recall, not keyword matching
const facts = await brain.ask("What's my job?");
// → ["I work at Google as a senior engineer"]

// Get LLM-ready context for any prompt
const context = await brain.context("coding help");
// → "Known facts about this user:\n- I work at Google..."

// Forget (GDPR right to be forgotten)
await brain.forget("I work at Google as a senior engineer");

Configuration

const brain = new Brain("alice", {
  apiKey: "your-api-key",             // or set CLS_API_KEY env var
  url: "https://www.clsplusplus.com", // or set CLS_BASE_URL env var
});

Full API

| Method | Description | |--------|-------------| | brain.learn(fact, meta?) | Teach a fact. Returns memory ID. | | brain.ask(question, limit?) | Query for relevant facts. Returns string array. | | brain.context(topic?, limit?) | Get LLM-ready context string. | | brain.forget(factOrId) | Forget by text or ID. Returns boolean. | | brain.absorb(content, source?) | Bulk-learn from text or array. Returns count. | | brain.who() | Auto-generated user profile. | | brain.correct(wrong, right) | Update a belief (forget old, learn new). | | brain.chat(message, llmFn?, system?) | Full conversation handler with memory. | | brain.teach(data) | Learn from key-value object. | | brain.watch(messages) | Learn from chat message array. | | brain.all(limit?) | Get all memories. | | brain.count() | Count stored memories. | | brain.wrap(llmFn) | Wrap any LLM function with auto-memory. |

Module-Level Functions

For scripts and one-liners:

import { learn, ask, context, forget } from "clsplusplus";

await learn("alice", "Prefers dark mode");
const facts = await ask("alice", "What theme?");

Use with OpenAI

import OpenAI from "openai";
import { Brain } from "clsplusplus";

const openai = new OpenAI();
const brain = new Brain("alice");

const response = await brain.chat(
  "Help me pick a framework",
  async (system, message) => {
    const res = await openai.chat.completions.create({
      model: "gpt-4",
      messages: [
        { role: "system", content: system },
        { role: "user", content: message },
      ],
    });
    return res.choices[0].message.content || "";
  },
);

Use with Anthropic

import Anthropic from "@anthropic-ai/sdk";
import { Brain } from "clsplusplus";

const anthropic = new Anthropic();
const brain = new Brain("alice");

const chat = brain.wrap(async (system, message) => {
  const res = await anthropic.messages.create({
    model: "claude-sonnet-4-20250514",
    max_tokens: 1024,
    system,
    messages: [{ role: "user", content: message }],
  });
  return res.content[0].type === "text" ? res.content[0].text : "";
});

const response = await chat("You are a helpful assistant", "What editor should I use?");

Browser Usage

The SDK uses native fetch and works in all modern browsers. Environment variables are not available in browsers, so pass apiKey and url directly:

const brain = new Brain("alice", {
  apiKey: "your-key",
  url: "https://www.clsplusplus.com",
});

Requirements

• Node.js 18 or later (uses native fetch)
• Zero runtime dependencies

License

Apache-2.0