Bridges AI assistants to Selenium WebDriver for full browser automation through MCP. You get the complete toolset: launch Chrome, Firefox, Safari, or Edge, navigate and interact with pages, find and click elements, type text, manage cookies, switch between windows, take screenshots, and execute JavaScript. Built with TypeScript and Zod validation, it exposes around 30+ discrete tools covering everything from basic navigation (browser_navigate, browser_click, browser_type) to advanced operations like drag and drop, file uploads, and wait strategies. Reach for this when you need Claude or another AI to drive actual browser sessions for testing, scraping, or automating complex web workflows that require JavaScript rendering and user interaction.
claude mcp add --transport stdio pshivapr-selenium-mcp -- npx -y selenium-webdriver-mcp