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

# Work together in Mantle

> Choose the right conversation, share AI context, and keep team decisions visible.

Mantle lets people and AI work in the same place. The goal is not only to get an AI answer, but to make the question,
context, review, and decision available to the people doing the work.

## A simple team workflow

<Steps>
  <Step title="Discuss the goal in a channel">
    Share the problem, owner, and deadline where the team can find them.
  </Step>

  <Step title="Bring in AI">
    Mention a model or agent in the channel, or open a shared AI chat when the work needs a longer back-and-forth.
  </Step>

  <Step title="Review together">
    Teammates can ask follow-up questions, challenge the answer, and add missing context.
  </Step>

  <Step title="Save the result">
    Put a lasting plan in a canvas, record the decision in the channel, and create any follow-up work in a connected
    app.
  </Step>
</Steps>

## Choose the right surface

* Use a [channel](/features/channels) for an open team discussion.
* Use a [thread](/features/messages-and-threads) for replies to one message.
* Use a [direct message](/features/direct-messages) for one-to-one conversation.
* Use a [shared AI chat](/features/ai-chats) for deep work that teammates may continue.
* Use a [canvas](/features/canvases) for a document that should outlive the conversation.
* Use a [huddle](/features/calls) when speaking is faster than typing.

## Make shared AI work easy to trust

* Include the source material or link to it.
* Say which parts are facts, suggestions, or open questions.
* Ask a person to review important legal, financial, security, or customer decisions.
* Keep the final decision beside the AI work instead of copying only the answer elsewhere.
* Remove sensitive content before creating a public share link.

<Tip>
  A useful handoff tells the next person what was asked, what the AI did, what has been checked, and what still needs a
  decision.
</Tip>


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