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

# Teams, connected tools, and history

> Deepmerge data model beyond memories: how teams, connected AI tools, context sprints, and revision history fit together.

Beyond [memories](/reference/memory), a few other pieces make up Deepmerge. You rarely
touch them directly; this is just so the words make sense when you see them.

## Team

Your **team** is the shared workspace itself, the boundary everything lives inside. You,
the people you invite, and all of your connected tools share one team. It is one shared
brain: everything saved by anyone's tool is visible to everyone on the team, with no
folders to set up. A finding saved by one tool is right there for the next.

## Connected tools

Each AI tool you connect (Claude, ChatGPT, Cursor, and so on) shows up as a **connected
tool**. Every memory records which tool wrote it, so the activity always answers "who
saved this?" without guessing.

## History

Every change is kept as **history**. When a tool edits or replaces a memory, the
previous version is preserved, along with which tool changed it and when. It works like
the version history in a shared document: the latest is up front, the full trail is
always there.

## Context Sprints

A **Context Sprint** is an interview campaign for one workflow. Deepmerge invites
participants, runs AI-guided conversations, writes a report from the evidence,
and saves reusable findings, insights, and skills back into the workspace.

Curious about the exact actions tools take behind the scenes? See the
[MCP tools reference](/reference/overview).