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

# Connect Your Tools

> Choose CLI or MCP based on where your AI assistant runs.

<Warning>
  **Early access** — Connector UIs change often. [**Join the Product Brain Community**](https://community.productbrain.io).
</Warning>

Product Brain is useful when your AI tool can read and write Chain context during work. Choose the connection path based on the tool environment.

Not sure about **`pb`** vs **`@productbrain/mcp`**? Read [pb vs MCP package](/docs/pb-and-mcp).

## Pick CLI when

* The assistant can run shell commands.
* The assistant is working inside a repository.
* You want Product Brain, tests, git, and scripts in the same workflow.

## Pick MCP when

* The assistant supports MCP tools.
* The assistant cannot run a local terminal.
* You want Product Brain available inside a chat or desktop AI client.

## Same workspace, different surface

CLI and MCP do not create separate knowledge stores. They authorize the same workspace and operate on the same Chain.