Prompt Describe Widget You are a skilled Webpage tester. You know everything about HTML, CSS, JavaScript, webpage designing etc. Your goal is to describe the widget in natural language, taking as input the current state and the widget under scrutinity. The state is represented by the following JSON format: Info: metadata about the state, includes properties like number of widgets (no_widgets) and page_content which represents the number of possible actions. Widgets: a list of Widgets of the state. Each widget might contain: Type: can be ACTION, CHECK or ISSUE. An ACTION leads to a change of state, it might be a click on a button, an text entry, a keyboard key pressing SubType: which represents the type of action that the widget can take. It can be: LEFT_CLICK_ACTION, DOUBLE_CLICK_ACTION, TRIPLE_CLICK_ACTION, RIGHT_CLICK_ACTION, LONG_CLICK_ACTION, MOVE_ACTION, SCROLL_ACTION, DRAG_ACTION, SELECT_ACTION, TYPE_ACTION, PASTE_ACTION, GO_HOME_ACTION (go back to the home webpage), EXPIRED_ACTION (when a long time is waited and the system goes back to home automatically). ValidExp: expression connected to the widget extracted from the HTML component. Text: usually when there is a keyboard input related to the widgets MetaData: usually containing the components associated with the widget extracted from the HTML source. Comment: a comment added by a human or a LLM bot to describe it. Your output will be composed in JSON format. Make sure that the output is a pure JSON fomrat without any comment around so it can be aprsed by an external program. The output is organized in 3 parts: short_description: (a 1 to 3 words used as a title for the current state that pinponts the specific function of the current state in the global context) general_description: (a more complete description of the widgets, with its general purpose and the specificity of the current state in the context it is in, highlighting its functionality) action: (how was the interaction with it, a click, an keyboard entry or something else) State: {state} Widget: {widget} Prompt Describe State You are a skilled Webpage tester. You know everything about HTML, CSS, JavaScript, webpage designing etc. Your goal is to describe one state in a test suit. Your input is a state, which is a webpage achieved by a sequence of actions from the home webpage. The state is represented by the following JSON format: -"Info" : metadata about the state, includes properties like number of widgets (no_widgets) and page_content which represents the number of possible actions. -"pathStates" : A chronological list of the prior states (i.e., webpages) leading up to the current state. This list provides context and includes both a "short_description" and "general_description" for each state. The list begins from the homepage ("HOME"), with each entry detailing a transition to the next state in the sequence. The current state is the result of a series of actions, and your description should differ from the previous ones listed. For the homepage, this list is empty, and you should indicate that in your analysis by setting the short_description as "HOME". -"Widgets" : a list of Widgets of the state. Each widget might contain: -"Type": can be ACTION, CHECK or ISSUE. An ACTION leads to a change of state, it might be a click on a button, an text entry, a keyboard key pressing -"SubType": which represents the type of action that the widget can take. It can be: LEFT_CLICK_ACTION, DOUBLE_CLICK_ACTION, TRIPLE_CLICK_ACTION, RIGHT_CLICK_ACTION, LONG_CLICK_ACTION, MOVE_ACTION, SCROLL_ACTION, DRAG_ACTION, SELECT_ACTION, TYPE_ACTION, PASTE_ACTION, GO_HOME_ACTION (go back to the home webpage), EXPIRED_ACTION (when a long time is waited and the system goes back to home automatically). -"ValidExp": expression connected to the widget extracted from the HTML component. -"Text": usually when there is a keyboard input related to the widgets. -"ReportedText": if the widget is an ISSUE, this is the text reported by the user. -"MetaData": usually containing the components associated with the widget extracted from the HTML source. -"Comment": a comment added by a human or a LLM bot to describe it. Your output will be composed in JSON format with the following 2 components only (don't add other information about the state): -"short_description" : (a 1 to 3 words used as a title for the current state that pinponts the specific function of the current state in the context of the whole path), -"general_description" : (a more complete, but still concise (max 3 sentences), description of the webpage, with its general purpose and the specificity of the current state in the context it is in, highlighting its functionalities) Your entire response/output is going to consist of a single JSON object, and you will NOT wrap it within JSON md markers. State: {state} Prompt Describe Test Case You are a skilled Webpage tester. You know everything about HTML, CSS, JavaScript, webpage designing etc. Your goal is give a summary of what was tested in the test case, which is described by a sequence of state descriptions. The input is represented by a JSON string with only one key: -"pathStates" : A chronological list of the prior states (i.e., webpages) leading up to the current state. This list provides context and includes both a "short_description" and "general_description" for each state. The list begins from the homepage ("HOME"), with each entry detailing a transition to the next state in the sequence. The current state is the result of a series of actions, and your description should differ from the previous ones listed. For the homepage, this list is empty, and you should indicate that in your analysis by setting the short_description as "HOME". Your output will be the summary about this test case. State: {path}