Welcome to the first episode of Quest to Compile 🎮

In this episode, we explore how GitHub Copilot + Unity MCP (Model Context Protocol) changes modern game development workflows. Instead of constantly switching between VS Code and Unity, you can have Copilot interact directly with the Unity Editor to create scenes, generate scripts, validate changes in real time, and more. 

We walk through setup, best practices, and a hands-on demo where we build a playable Unity prototype from prompts using Unity, .NET, and GitHub Copilot. 

Along the way, expect a bit of sibling banter--because building games runs in the family.

🚀 What You’ll Learn
How to connect GitHub Copilot to Unity using MCP
What Model Context Protocol (MCP) is and why it matters
How AI can create and modify Unity scenes in real time
How to use Copilot instructions, custom agents, and skills
How to debug AI workflows (permissions, disconnects, domain reloads)
How to prototype gameplay quickly using AI-assisted development


⏱️ Chapters
00:00 Intro: Stop Alt-Tabbing Between VS Code and Unity
00:23 What Unity MCP + Copilot Can Do
01:57 Installing Unity MCP (AI Assistant Package)
04:02 Understanding MCP: Bridge, Relay, and Tools
06:02 What Is Model Context Protocol (MCP)?
08:16 Tool Permissions and Execution Model
09:00 Setting Up MCP in VS Code
10:17 Registering the Unity MCP Server
15:00 Best Practices: Instructions, Agents, and Skills
16:06 Auto-Generating Project Context with MCP
17:36 Custom Agents for Game Development
19:56 Skills vs Agents Explained
23:06 Demo: Building a Scene with AI Prompts
36:42 Key Takeaways and Best Practices
37:10 Extending MCP with Custom Tools

🔗 Links & Resources
Learn about GitHub Copilot: https://github.com/features/copilot
Unity MCP documentation: https://on.unity.com/AIDocumentation
Visual Studio Code: https://code.visualstudio.com/
📺 Watch the full series: aka.ms/Quest-to-Compile


🏷️ Hashtags
#GameDevelopment #Unity3D #DotNet #GitHubCopilot #AIDevelopment
#VSCode #IndieDev #gamedev #GameDevTools #Programming #SoftwareEngineering
#AIForDevelopers #BuildInPublic #DevWorkflow #CSharp #UnityDev

Microsoft Developer

A walkthrough of using GitHub Copilot with the Unity MCP (Model Context Protocol) server to build a simple Unity game without leaving VS Code. The Unity MCP server lets Copilot interact directly with the Unity editor for scene management, material creation, profiler access, and more. Setup involves installing the Unity AI Assistant package, configuring the MCP relay in VS Code, and optionally adding Copilot instruction files, custom agents, and skills for better results. A live demo shows Copilot creating a ground plane, coloring objects, placing collectible coins, writing player movement scripts, and adding coin collection logic — all through natural language prompts. Custom MCP tools like domain reload triggering, game-camera screenshots, and input injection are also discussed as extensions.

Build a Unity Game with GitHub Copilot + MCP | .NET Game Development