{"slug":"playwright","url":"https://agentics.page/skills/playwright","name":"Playwright Browser","vendor":"Microsoft","official":true,"category":"browser-automation","categoryTitle":"Browser automation","summary":"Lets your assistant open and operate a real web browser.","detail":"Playwright MCP gives an assistant full control of a real browser: navigate to pages, click buttons, fill forms, read what is on screen, and take screenshots. It is the skill behind use cases like \"compare prices across three shops\" or \"fill in this web form for me\". Because it drives an actual browser, it works on sites that plain fetching cannot read.","install":{"kind":"npx","package":"@playwright/mcp@latest"},"envKeys":[],"repoUrl":"https://github.com/microsoft/playwright-mcp","worksWith":["claude-desktop","claude-code","cursor","vscode"],"recipes":[{"title":"Let your AI assistant browse the web and compare prices for you","url":"https://agentics.page/recipes/let-ai-browse-the-web-for-you","api":"https://agentics.page/api/recipes/let-ai-browse-the-web-for-you"}]}