Playwright MCP Server

# API Reference ## PlaywrightMCPServer Main server class that implements the MCP protocol for Playwright browser automation. ### Methods #### `__init__()` Initialize the Playwright MCP Server. #### `run()` Start the MCP server and begin listening for requests. #### `cleanup()` Clean up browser resources and stop the server. ## PlaywrightTools Collection of browser automation tools available through the MCP server. ### Navigation Methods #### `navigate(page: Page, url: str) -> str` Navigate to a specific URL. **Parameters:** - `page`: Playwright page instance - `url`: Target URL to navigate to **Returns:** - Success/error message string #### `get_url(page: Page) -> str` Get the current page URL. #### `get_title(page: Page) -> str` Get the current page title. ### Element Interaction Methods #### `click(page: Page, selector: str, timeout: int = 5000) -> str` Click on an element matching the selector. **Parameters:** - `page`: Playwright page instance - `selector`: CSS selector for the target element - `timeout`: Maximum wait time in milliseconds #### `type_text(page: Page, selector: str, text: str, timeout: int = 5000) -> str` Type text into an input element. **Parameters:** - `page`: Playwright page instance - `selector`: CSS selector for the input element - `text`: Text to type - `timeout`: Maximum wait time in milliseconds #### `select_option(page: Page, selector: str, value: str, timeout: int = 5000) -> str` Select an option from a dropdown. ### Content Extraction Methods #### `get_text(page: Page, selector: str, timeout: int = 5000) -> str` Extract text content from an element. #### `get_attribute(page: Page, selector: str, attribute: str, timeout: int = 5000) -> str` Get an attribute value from an element. #### `screenshot(page: Page, path: Optional[str] = None, full_page: bool = False) -> str` Take a screenshot of the page. ### Wait Methods #### `wait_for_selector(page: Page, selector: str, timeout: int = 30000, state: str = "visible") -> str` Wait for an element to reach a specific state. **Parameters:** - `state`: One of "visible", "hidden", "attached", "detached" #### `wait_for_load_state(page: Page, state: str = "load") -> str` Wait for a specific load state. **Parameters:** - `state`: One of "load", "domcontentloaded", "networkidle" ### JavaScript Methods #### `evaluate(page: Page, script: str) -> str` Execute JavaScript code in the browser context. ## Error Handling All methods return string results that indicate success or failure. Errors are caught and returned as descriptive error messages rather than raising exceptions. ## Browser Management The server automatically manages: - Browser instance lifecycle - Page/tab creation and cleanup - Context isolation - Resource cleanup on shutdown