Context & OCR
Context makes prompt turns precise. Attach your selection, active window, clipboard, or drag and drop files onto the composer card.
Context attachments
Ruru provides dedicated attachment triggers directly inside the prompt composer:
- Selected Text: Automatically captures highlighted lines from your active editor or document.
- Active Window: Captures the frontmost window and titles to give the model awareness of your current task.
- Entire Screen: Captures your display to analyze full layouts or multi-window setups.
- Clipboard: Pulls the latest copied text snippet into the turn context.
Drag & drop onto the composer
Drop files of any kind — images, PDFs, source code files, or data spreadsheets — anywhere on the Ruru composer card.
- Files join the current turn as removable context pills.
- Any prompt text you were already in the middle of typing remains completely intact.
- Click on an attached context pill to preview it with macOS QuickLook before submitting.
On-device OCR with Apple Vision
When you capture an active window or screen region containing non-selectable text (such as an Xcode canvas, simulator window, or video frame), Ruru extracts text using macOS Apple Vision framework.