> ## Documentation Index
> Fetch the complete documentation index at: https://docs.xysq.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Introduction

> xysq is a consent-first memory layer for AI agents.

## What is xysq?

xysq gives AI agents persistent, identity-aware memory — with you in control of what they remember.

When you talk to an AI agent today, it forgets everything the moment the conversation ends. xysq changes that. It gives agents a place to store and retrieve context across sessions, tools, and platforms — without you having to repeat yourself every time.

The difference is consent. xysq is built around the idea that your memory belongs to you. Agents can only retain and recall information within the boundaries you set.

## How it works

xysq connects to your AI agents via the **Model Context Protocol (MCP)** — a standard interface that tools like Claude and Cursor support natively.

Once connected, agents can:

* **Retain** — store something you want remembered
* **Recall** — retrieve relevant memories from past sessions
* **List** — browse what's stored
* **Delete** — remove anything you no longer want an agent to know

Memory is scoped to your identity. Whether you're using Claude, Cursor, or any other MCP-compatible tool, it's the same memory.

## Get connected

<CardGroup cols={2}>
  <Card title="Connect your agent" icon="plug" href="/integrations/connect-agent">
    Connect xysq to Claude, Cursor, or any MCP agent in a few steps.
  </Card>

  <Card title="Install the skill" icon="download" href="/features/install-guides">
    Teach your agent how to use xysq from the first session.
  </Card>
</CardGroup>

## Quick example

Once connected, it's as natural as asking:

```
Remember that I prefer TypeScript over JavaScript in all new projects.
```

```
What do you know about my preferences?
```

xysq handles the storage and retrieval — your agent just uses it.