Docking documents with chat

The AI chat doesn't need to "attach" to a document. Whenever you're on a document page, the assistant already knows which document you're on — title, your selection, your cursor, and the section headings currently in view get sent with every message. What you do control is where the chat itself sits on the screen: as a floating window in the corner, or as a panel docked to the right of the page.

This page covers how to open it, the two display modes, what the assistant actually sees from the document, and where its awareness ends.

Opening the chat

The chat opens from the round gradient logo button that sits in a corner of the viewport. It's draggable — grab it and drop it in any of the four corners; the position persists.

There is no keyboard shortcut to open the chat. The closest thing is the command palette: press ⌘K, type an action-style query ("summarise this document", "draft a methods paragraph"), and press Enter (or Tab to force AI mode, then Enter). The palette routes the query to the chat and opens it for you. See Command palette.

Floating window vs. docked side panel

The chat panel itself has two display modes. The toggle is the small window/sidebar icon in the chat tab bar header.

Floating window

A 420-pixel panel anchored to a corner. Draggable. Good for quick questions and short replies — it overlays the page, so it can cover content.

Docked side panel

A full-height column glued to the right edge of the page. The document reflows to make room, so editor and chat sit truly side-by-side. Good for sustained editing while chatting.

Switching modes preserves the conversation — and crucially, your document context (title, selection, visible sections) is sent on every message in both modes. Docking the chat is a layout preference, not a way of "giving the assistant access" to the document.

What the assistant sees from the document

When you're on a document page, every message you send carries a small context payload alongside the text you typed:

• Document title and ID — so the assistant can name it back to you.
• Your selection or cursor position — if you've highlighted text, that exact range is in the payload; if not, the cursor's location is.
• Visible section headings — a list of the headings currently scrolled into view, so the assistant has a rough sense of where in the document you are.

That's the whole picture. A few things the assistant does not get:

• A full block-by-block outline of the document.
• Block IDs it can cite. References like "the protocol step about plasma collection" work only if that phrase is in a heading or in your visible selection.
• A diff of what you just changed. The assistant sees the document's current state at the moment you send the message — not what you edited thirty seconds ago. If you want it to react to a specific edit, tell it about the edit in your message.

In practice this last point matters most. If you've just rewritten a paragraph and want feedback, don't ask "is that better?" — say "I just rewrote the second paragraph in the methods section; does it follow from the table above?" and you'll get a useful answer.

Patterns that work well

Drafting from visible content

Scroll the section you want the assistant to use into view, then ask "draft a methods paragraph from the steps in view". Visible-section context tells the assistant where to look.

Working on the selection

Highlight the paragraph you want feedback on, then ask. The exact selection is sent — no need to paste it.

Calling out a recent edit

Because the chat doesn't track changes, mention edits explicitly: "I just changed the assay buffer from PBS to TBS — does the rest of the protocol still hold?".

Comparing two documents

Open document A, ask the assistant for a summary. Switch to document B, ask "how does this differ from what you just summarised?". The conversation persists across pages.

Attaching images and screenshots

On a vision-capable model — Claude 4.x, GPT-5.x, Gemini 2.5+ / 3.x, Gemma 4, and the Mistral / Magistral family — you can drag an instrument photo, a gel scan, or a UI screenshot straight into the chat composer (or paste it with ⌘V). The image sits inline above your prompt and the model sees it on the next turn.

Useful pairings with a document open:

• Annotate a chart. With the document containing the chart in view, drop a screenshot and ask "where on the chart does this anomaly come from?".
• Extract from an instrument capture. Drop a plate-reader screenshot and ask the assistant to transcribe the values into the result table you have open.
• Reproduce a layout. Drop a screenshot of a published figure and ask "rebuild this as a chart block in the document I have open".

If your current model is text-only (gpt-oss 120b, DeepSeek V4) the paperclip icon is disabled and image drops are ignored — switch to a vision model from the chat header to enable attachments.

What's next