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

# What is Goose?

> An AI agent that lives in your workspace and does real work

**Goose** is the AI agent at the heart of Gooseworks. It's not a chatbot that forgets everything after each conversation — it's a persistent coworker that operates from its own workspace with files, email, memory, and tools.

## What makes Goose different?

| Feature | Typical AI chat | Goose |
| - | - | - |
| **Memory** | Forgets after each session | Remembers context across sessions |
| **Files** | No file system | Has its own cloud filesystem |
| **Email** | Can't send email | Has its own inbox and can send/receive |
| **Tools** | Limited or none | Extensible via skills and MCP servers |
| **Automation** | Manual only | Can run scheduled jobs on its own |

## How you interact with Goose

Your primary interface is the **chat**. You type what you need, and Goose figures out how to get it done — using its skills, files, research tools, and integrations.

Goose can:

* Stream responses in real time as it works
* Show you interactive widgets like data tables, email drafts, and action buttons
* Ask you for clarification when it needs more context
* Create and modify files in your workspace
* Draft and send emails on your behalf

## Goose is workspace-aware

Everything Goose does happens within your **workspace**. Your files, chat history, skills, environment variables, and memory all persist — so the next time you chat, Goose already knows your context.

This means work **compounds**. The research Goose does today is available tomorrow. The skills you build this week save you time next week.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.