MCP Browser Kit

🌐 GET BROWSER CONTEXT - CRITICAL FIRST STEP BEFORE USING ANY OTHER TOOLS! * This tool MUST be called first to initialize browser automation and get essential data. * Returns data structure with: - tabs: Array of browser tabs with properties like id, url, title, and active status - manifestVersion: Version of extension manifest format supported by the browser * Each tab includes a unique tabKey required for all other tool operations * The active tab (marked with 'active: true') is typically your target for automation * The manifestVersion determines which browser features and extension capabilities are available * Different browsers support different manifest versions, affecting available tools and API access * Standard workflow: 1) getBasicBrowserContext → get browser state and tabKey 2) Analyze page content based on your goal and manifest version: - If interaction is required (clicking, filling forms, etc.): · For Manifest Version 2: Use captureTab for visual context or getReadableElements for element identification · For other Manifest Versions: Use only getReadableElements for element identification - If no interaction is required (just reading page content): · Use getReadableText to extract all visible text from the page 3) Interact using click/fill/enter tools with the obtained tabKey

📷 Captures a screenshot of a browser tab * Use this tool after calling getBasicBrowserContext to obtain visual context of the page * The screenshot helps you see what the browser is displaying to the user * Requires tabKey from getBasicBrowserContext * Returns an image with width, height, and data in base64 format * Workflow: 1) getBasicBrowserContext → 2) captureTab → 3) interact with elements * Parameters: tabKey * NOTE: This feature is only available in browsers supporting Manifest Version 2

⚙️ Executes custom JavaScript code in the context of the web page * Use this for advanced operations not covered by other tools * Requires tabKey from getBasicBrowserContext and JavaScript code to execute * The code should be the body of a function that returns a value * Example: 'return document.title;' to get the page title * Gives you full flexibility for custom browser automation * Parameters: tabKey, fnBodyCode (JavaScript code as string) * NOTE: This feature is only available in browsers supporting Manifest Version 2

🌐 Opens a new browser tab with the specified URL * Use this to navigate to a new page in a new tab * Requires windowKey from getBasicBrowserContext and the URL to open * Returns the tabKey of the newly created tab which you can use for further operations * The new tab will be created in the specified browser window * After opening, you may need to wait a moment for the page to load * Parameters: windowKey, url

🗑️ Closes a specific browser tab * Use this to close a tab when you're done with it or need to clean up * Requires tabKey from getBasicBrowserContext * The tab will be permanently closed and cannot be recovered * Be careful not to close the tab you're currently working with * Parameters: tabKey