Add context to an AI request
Attach selected text, files, links, sources and a writing style to your next message.
Use Add content (the plus button) in the AI message box when a request needs material beyond the open file. You can also type @ in the message box to find context by name.
Choose material
The menu offers these choices:
| Choice | What it adds |
|---|---|
| Current Document | Text from the open Quill document. In a presentation, this row reads Current presentation and adds a slide outline. |
| Upload files | A file from your device. |
| Add link | Readable text fetched from the URL you enter. |
| Documents | A document you can access in Sourcelett. |
| Files | One of your saved files. |
| Sources | A saved research source. |
| Writing style | A saved style for how the AI writes this response. |
Choose a category, use Search... to find an item, then select it. The item appears as an attachment above the message box. Select its × to remove it before sending. You can attach more than one item, but selecting another writing style replaces the previous style for that message.
The attachments go with the message you send. The composer clears the chips afterward, while that conversation keeps the material as context. Attach it again when you start a new chat.
Use a selection or file from your device
In a Quill document, select a passage and choose Add to chat on the selection toolbar. The AI panel opens with Selected Text attached. Ask about that passage or switch to Agent mode to request an edit.
For a local file, choose Upload files, drag it into the composer, or paste an image. Each file must be 10 MB or smaller. PDF, DOCX and PPTX files are converted to text before the AI receives them. PNG, JPEG, WebP and GIF images require a model with image support; choose one in the model menu before attaching the image. A file read or conversion error appears above the message box.
Was this helpful?