Screen recording
Lace uses screen recording permission to capture what’s visible on your display. This is how it builds screen context: taking snapshots of your active window so the AI can see what you’re working on. macOS will prompt you when screen context is first enabled. You can also grant it later in System Settings > Privacy & Security > Screen Recording. Without this permission, Lace can still chat but won’t reference anything on your screen.Accessibility
Lace reads the macOS accessibility tree to understand the structured content behind what’s visible: text labels, button names, input values, and UI element roles. This gives Lace a richer understanding than screenshots alone. Combined with vision, accessibility data lets Lace:- Read text content in any app (not just browsers)
- Identify interactive elements like buttons, links, and inputs
- Understand page structure and navigation hierarchy
Microphone
Lace uses the microphone to transcribe voice input in chat. Press ⌘D or tap the mic button in the chat input to start dictating. Your speech is transcribed and inserted into the message field. macOS will prompt you when you first use dictation. You can also grant it in System Settings > Privacy & Security > Microphone. Without this permission, the dictation feature is unavailable but all other functionality works normally.Why all three?
| Permission | What it sees | Why it matters |
|---|---|---|
| Screen recording | Pixels on screen | Visual layout, images, charts, spatial relationships |
| Accessibility | UI element tree | Text content, element roles, interactivity, structure |
| Microphone | Audio input | Voice-to-text dictation in chat |