> ## Documentation Index
> Fetch the complete documentation index at: https://docs.spongecake.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Desktop Actions

> Below are desktop actions that can be performed by the Desktop class.

### Functions

<AccordionGroup>
  <Accordion icon="message-bot" title="click(x, y, click_type='left')">
    ```python theme={null}
    def click(self, x: int, y: int, click_type: str = "left") -> None:
        """
        Move the mouse to (x, y) and click the specified button.
        click_type can be 'left', 'middle', or 'right'.
        """
    ```

    **Arguments**:

    * **x, y** *(int)*: The screen coordinates to move the mouse.
    * **click\_type** *(str)*: The mouse button to click (`"left"`, `"middle"`, or `"right"`).

    **Returns**:

    * `None`
  </Accordion>

  <Accordion icon="message-bot" title="scroll(x, y, scroll_x=0, scroll_y=0)">
    ```python theme={null}
    def scroll(
        self,
        x: int,
        y: int,
        scroll_x: int = 0,
        scroll_y: int = 0
    ) -> None:
        """
        Move to (x, y) and scroll horizontally or vertically.
        """
    ```

    **Arguments**:

    * **x, y** *(int)*: The screen coordinates to move the mouse.
    * **scroll\_x** *(int)*: Horizontal scroll offset.
    * Negative => Scroll left (button 6)
    * Positive => Scroll right (button 7)
    * **scroll\_y** *(int)*: Vertical scroll offset.
    * Negative => Scroll up (button 4)
    * Positive => Scroll down (button 5)

    **Behavior**:

    * Moves the mouse to `(x, y)`.
    * Scrolls by scroll\_x and scroll\_y

    **Returns**:

    * `None`
  </Accordion>

  <Accordion icon="message-bot" title="keypress(keys: list[str])">
    ```python theme={null}
    def keypress(self, keys: list[str]) -> None:
        """
        Press (and possibly hold) keys in sequence.
        """
    ```

    **Arguments**:

    * **keys** *(list\[str])*: A list of keys to press. Example: `["CTRL", "f"]` for Ctrl+F.

    **Behavior**:

    * Executes a keypress
    * Supports shortcuts like Ctrl+Fs

    **Returns**:

    * `None`
  </Accordion>

  <Accordion icon="message-bot" title="type_text(text: str)">
    ```python theme={null}
    def type_text(self, text: str) -> None:
        """
        Type a string of text (like using a keyboard) at the current cursor location.
        """
    ```

    **Arguments**:

    * **text** *(str)*: The string of text to type.

    **Behavior**:

    * Types a string of text at the current cursor location.

    **Returns**:

    * `None`
  </Accordion>

  <Accordion icon="message-bot" title="get_screenshot()">
    ```python theme={null}
    def get_screenshot(self) -> str:
        """
        Takes a screenshot of the current desktop.
        Returns the base64-encoded PNG screenshot.
        """
    ```

    **Behavior**:

    * Takes a screenshot as a png
    * Captures the base64 result.
    * Returns that base64 string.

    **Returns**:

    * *(str)*: A base64-encoded PNG screenshot.

    **Exceptions**:

    * **RuntimeError** if the screenshot command fails.
  </Accordion>

  <Accordion icon="message-bot" title="goto(url)">
    ```python theme={null}
    def goto(self, url: str) -> None:
        """
        Open Firefox in the container and navigate to the specified URL in a new tab.
        """
    ```

    **Arguments**:

    * **url** *(str)*: The URL to navigate to (e.g., `https://example.com`).

    **Behavior**:

    * Opens Firefox in the container and navigates to the specified URL in a new tab

    **Returns**:

    * `None`
  </Accordion>

  <Accordion icon="message-bot" title="get_page_html(query)">
    ```python theme={null}
    def get_page_html(self, query="return document.documentElement.outerHTML;"):
        """
        Get the HTML content of the currently displayed webpage using Marionette.
        
        Args:
            query: JavaScript query to retrieve the HTML content. Defaults to retrieving the full DOM HTML.
        
        Returns:
            str: The HTML content of the current page, or an error message if retrieval fails.
        """
    ```

    **Arguments**:

    * **query** *(str)*: JavaScript query to retrieve the HTML content.
      Defaults to `'return document.documentElement.outerHTML;'`, which returns the full DOM HTML.

    **Behavior**:

    * Uses the underlying agent to execute a JavaScript query and retrieve the current webpage's HTML content.

    **Returns**:

    * *(str)*: The HTML content of the current page, or an error message if retrieval fails.
  </Accordion>
</AccordionGroup>