🤖 AI Development Assistant MCP Server

What is AI Development Assistant MCP Server?

AI Development Assistant MCP Server is an MCP server designed for Cursor that provides intelligent coding assistance through custom AI tools. It includes tools for code architecture planning, UI screenshot analysis, and git-based code review, but is primarily a tutorial demo and not production-ready.

How to use AI Development Assistant MCP Server?

Install dependencies with npm install (or yarn install), set your OpenAI API key in src/env/keys.ts, then build the server with npm run build-only. After building, add it to Cursor by going to Cursor Settings > Features > MCP, clicking + Add New MCP Server, setting Type to stdio, and entering node /path/to/build/index.js as the Command.

Key features of AI Development Assistant MCP Server

• Code Architect: calls advanced reasoning LLMs to generate plans and instructions
• Screenshot Buddy: takes UI design screenshots for use with the composer agent
• Code Review: uses git diffs to trigger code reviews
• Local development only; API keys stored in source file

Use cases of AI Development Assistant MCP Server

• Automating code architecture generation for new features
• Analyzing UI screenshots to improve designs
• Performing code reviews based on git diffs
• Learning how to build and integrate an MCP server in Cursor

FAQ from AI Development Assistant MCP Server

Is this server production-ready?

No. The README explicitly states it is mostly a tutorial demo and not a production-ready tool.

What API key is required?

An OpenAI API key must be set in src/env/keys.ts. The README notes this is only for local development and learning.

How is the server connected to Cursor?

It uses the stdio transport type. You add it as a new MCP server in Cursor Settings with a command pointing to the built index.js.

Where do the tools appear after setup?

The tools (Code Architect, Screenshot Buddy, Code Review) appear under "Available Tools" in Cursor's MCP server section, and can be used in Cursor's Composer.

What are the known limitations?

The project is intended as a tutorial demo. The author states they likely won't fix issues, and recommends forking the project for further development.