Canvas documents overview
Rich-text pages for notes, specs, and decisions that live alongside the conversation.

A canvas document is a rich-text page that lives inside a space, right next to the channel. Use one for anything that wants more structure than a chat message: meeting notes, runbooks, specs, design docs, decision records.
What makes them different from a Google Doc
- They live with the conversation. Open the space, the document is in the same sidebar as the channel and any agents. You don't switch apps to read it.
- They're collaborative in real time. Several people (and agents) can edit at the same time. You see live cursors, and edits merge cleanly without a "someone else's changes" overlay.
- They support comments anywhere. Highlight any text — or pick a whole block from the drag handle — and start a thread. Replies show up next to the document and feed into the same notification routing as messages.
- AI agents can read and edit them too.
@mentionan agent in a message and ask it to "summarise the doc" or "rewrite the intro" — it edits the same canvas everyone else can see, with its contributions colour-coded so you can tell who wrote what.
What's covered in this section
- Editing a document — every block type, mark, menu, slash command, and keyboard shortcut.
- Authorship colours & inline autocomplete — see who wrote which sentence, and let an AI finish your thoughts.
- Mentions —
@-mention people and agents inside the document;#-mention spaces, other canvases, and tasks. - Version history — save checkpoints, view and compare earlier versions, and revert when you need to.
Mentioning a document in a message
You can #mention any canvas document from a message. The mention
turns into a chip everyone in the space can click to jump to the
document, and the chip's label stays in sync as the document is
renamed.
You can also #mention documents, databases, spaces, and tasks from
inside a canvas document — see Mentions for the full
reference.
Permissions
Documents inherit access from the space they live in. Anyone who can read the space can read and edit its documents. There's no separate per-document permission to learn.
