MCP Servers are trending, but it’s not always clear how to use them effectively to boost Cursor’s productivity.
Here are five key aspects about MCP servers that had the greates impact:
- Playwright automation for frontend testing - A must-have tool for frontend development. Playwright lets AI run UI tests and validate components, creating a seamless loop between coding and testing that fully automates the process.
- Browser agent for research and complex tasks - Cursor can perform complex web tasks like checking documentation, conducting research, and testing compex scenarios for services I develop by connecting browser-use as an MCP server.
- Describe tools with .mdc rules - Adding descriptions of when to use each tool improved AI’s decision-making. It now selects the right tool without me having to specify or force Cursor to use a particular option.
- Dev mode for MCP servers development - Using the SDK dramatically simplifies MCP server development. You can test tools, resources, and see all issues directly from the UI.
- Use Cursor beyond coding - I now use Cursor as an AI-driven editor for database tasks, GitHub issue tracking, research, and maintaining notebooks - all through MCP server connections.
My previous practical tutorial received positive feedback so I wrapped everything about MCP servers into an interactive format as well: https://enlightby.ai/projects/11
There you’ll find a step-by-step guide to setting up MCP servers, learning how to develop and test them, and building your own AI-agentic browser tool right in Cursor. It’s completely free.
Also available as an extension: Enlighter for Cursor – Learn Vibe Coding - Visual Studio Marketplace
What MCP server do you use with Cursor? What are the most useful MCP servers you’ve found or developed?