Screenshots on a computer display are automatically captured to assist in building a tutorial to demonstrate the functionality of one or more applications. In one possible approach, an event detector, such as an accessibility toolkit, detects when events are generated by the one or more applications. The events may include, e.g., button clicks, windows opening or closing, drag and drop events, check boxes being checked, menus being pulled down or menu items selected, and so forth. Or, a program may monitor pixel changes on the screen to detect when changes occur in the images generated by the one or more applications. A tutorial builder application automatically captures an image when the event or image change is detected, edits the captured image such as by providing a focus on the changed portion, and provides the edited image in a tutorial window. Related text can also be automatically generated.