The Sage_MacOS project provides a macOS application which allows starting a Sage session running either in a command line or in a Jupyter or JupyterLab notebook. It is distributed as a single file disk image which can be installed by the familiar process of dragging the application icon to the Applications folder. The application and the disk image are signed and notarized, so no special set up is required to run the app.
When the app is started by clicking its icon it opens a small window which allows selecting a user interface. Clicking the "Launch" button starts Sage running with the chosen command line or notebook interface and closes the window. If a notebook interface was requested, a Jupyter server will be started. The window will reopen if the icon is clicked again, allowing starting a new session.
As with a typical macOS application, SageMath continues to run until it is stopped by selecting Quit from the File menu, or typing Command-Q, or selecting Quit from the contextual menu on the Dock icon. When the app quits it will terminate the Jupyter server process, if one was started. (This is a change starting with version 10.3, meant to prevent having unused server processes running on the system.)
The disk image includes an Installer package which creates an executable script /usr/local/bin/sage that can be used to start sage from a shell or script. It also installs Jupyter kernel specifications in a standard location in /usr/local to enable other Jupyter applications to use the Sage kernel.
The sage script does not support the -i option, because installing new packages inside of a notarized application bundle causes the signature to become invalid and the application to be unusable. However, the application already includes as many optional packages as possible. In addition, the Sage %pip command can be used to install PyPI packages in the user's .sage directory, allowing them to be accessed during a Sage session.
