jstricklin/Unity-Intelligence-MCP
If you are the rightful owner of Unity-Intelligence-MCP and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.
The Unity Code Intelligence MCP Server is an AI-powered static analysis tool designed to enhance Unity project development by detecting patterns, mapping component relationships, and providing insightful metrics.
Unity Code Intelligence MCP Server
AI-powered static analysis for Unity projects
This Model Context Protocol (MCP) server provides intelligent code analysis for Unity, detecting common patterns, component relationships, and programming insights to augment developer workflows.
Key Features
- 🎮 Unity-Specific Pattern Detection - Identifies 8+ common Unity patterns (Singleton, Coroutine, ScriptableObject etc.)
- 🤝 Component Relationship Mapping - Visualizes MonoBehaviour dependencies and interactions
- 📊 Quantitative Metrics - Provides statistics on pattern usage across projects
- 🔍 Structured Documentation Access - AI-optimized Unity documentation retrieval
- ⚙️ Automatic Configuration - Resolves Unity installations based on project settings
Designed for Extensibility
Modular Pattern Detection
Our pattern detection system uses a plugin architecture where new Unity patterns can be detected by implementing the IUnityPatternDetector
interface. The registry automatically discovers and integrates all detectors found in the system.
Flexible Configuration
The dual-path configuration system supports both automatic Unity installation detection and manual overrides. This adaptive approach ensures compatibility with future Unity versions and custom deployment scenarios.
Documentation Framework
The documentation engine provides a clean abstraction layer over Unity's documentation, offering multiple endpoints optimized for AI consumption. New documentation formats and search strategies can be added via our standardized plugin API.
Getting Started
Prerequisites
Before you begin, ensure you have the .NET 8.0 SDK or a newer version installed.
Building and Running
The following instructions apply to both Windows and macOS.
1. Build the Server
From the root directory of the repository, run the following command to build the project and restore dependencies:
dotnet build
2. Run the Server
To start the server, execute the following command from the root directory:
dotnet run
The server will now be running and ready to accept requests from an MCP client.
Configuration
The server is configured using the appsettings.json
file, which must be placed in the project's root directory. This file allows you to specify paths required for the analysis tools to function correctly.
UnityAnalysisSettings
All settings are located within the UnityAnalysisSettings
object in appsettings.json
.
-
ProjectPath
(Required): The absolute path to the root of the Unity project you intend to analyze.- Example:
"C:\\Users\\YourUser\\Documents\\MyUnityProject"
- Example:
-
InstallRoot
(Optional): The path to the directory containing your Unity Editor version folders. This is typically theEditor
folder inside your Unity Hub installation directory. This setting is used with the project's detected version to automatically locate the correct Unity Editor.- Example:
"C:\\Program Files\\Unity\\Hub\\Editor"
- Example:
-
EditorPath
(Optional): A direct and explicit path to a specific Unity Editor installation folder. If provided, this path takes priority over theInstallRoot
setting. This is useful for development or if your editor is in a non-standard location.- Example:
"C:\\UnityEditors\\2022.3.15f1"
- Example:
To function, the analyzer must be able to locate the Unity Editor installation. You must configure either InstallRoot
(so the editor can be found automatically) or provide a direct EditorPath
.
Sample appsettings.json:
{
"UnityAnalysisSettings": {
"InstallRoot": "/Applications/Unity/Hub/Editor/",
"EditorPath": "",
"ProjectPath": "C:\\Path\\To\\Your\\UnityProject"
}
}
Interacting with the Server
You can interact with the server using the MCP Inspector, which provides a web-based interface.
# Navigate to your MCP server project directory
# (The directory containing Unity_Intelligence_MCP.csproj)
cd path/to/Unity_Intelligence_MCP
# Install and run the MCP Inspector
npx @modelcontextprotocol/inspector dotnet run
Follow the instructions in your terminal. You will be prompted to open a URL in your web browser to access the inspector.
From the web interface, you can select a tool like analyze_unity_project
. The tool will use the ProjectPath
configured in your appsettings.json
to run the analysis.