Unity MCP Server
A Unity Editor plugin that exposes Unity Editor capabilities to external AI agents via the Model Context Protocol. It enables AI agents to interact with Unity through tools for debugging, scene inspection, project management, and build operations.
README
Unity MCP Server
中文 | English
Unity Editor 插件,通过 MCP (Model Context Protocol) 将 Unity Editor 能力暴露给外部 AI Agent。
Agent(如 Kiro、Cursor、Claude Desktop)可通过标准 MCP 协议连接到 Unity Editor,调用编辑器功能。
特性
- Streamable HTTP 传输 — 基于 MCP 2025-03-26 规范,单一 HTTP 端点
- 进程内运行 — 无需外部 Node.js/Python 进程,直接在 Editor 内启动
- 可扩展工具系统 — 实现
IMcpTool接口即可注册新工具,零修改核心代码 - Domain Reload 自动恢复 — 进入/退出 PlayMode 后服务自动重启
内置工具
| 工具 | 分类 | 功能 |
|---|---|---|
console_getLogs |
debug | 获取 Unity Console 最近 N 条日志 |
debug_getStackTrace |
debug | 获取最近一条 Error/Exception 的完整堆栈 |
debug_getPerformanceStats |
debug | 获取 FPS、DrawCall、内存占用等性能指标 |
debug_screenshot |
debug | 截取 Game/Scene 视图截图(base64 PNG) |
menu_execute |
editor | 按路径执行 Unity 菜单项 |
playmode_control |
editor | 进入/退出/查询 PlayMode 状态 |
editor_getSelection |
editor | 获取当前选中的 GameObject 和 Asset 信息 |
editor_getHierarchy |
editor | 获取场景 GameObject 树结构(可限深度) |
editor_getProjectStructure |
editor | 获取 Assets 目录结构(可限深度) |
editor_getInspector |
editor | 获取选中对象的 Inspector 序列化字段值 |
build_compile |
build | 触发脚本编译并返回结果 |
build_getCompileErrors |
build | 获取当前编译错误列表 |
build_runTests |
build | 运行 Unity Test Runner 测试并返回结果 |
安装
Git URL 安装(推荐)
- Unity Editor → Window → Package Manager →
+→ Add package from git URL - 输入以下 URL:
https://github.com/yangfch3/Unity-MCP-Server.git
或直接编辑宿主项目的 Packages/manifest.json:
{
"dependencies": {
"com.yangfch3.unity-mcp": "https://github.com/yangfch3/Unity-MCP-Server.git"
}
}
本地路径安装
- 克隆本仓库
- Unity Editor → Window → Package Manager →
+→ Add package from disk - 选择本仓库根目录的
package.json
或在 Packages/manifest.json 中添加:
{
"dependencies": {
"com.yangfch3.unity-mcp": "file:../../path/to/unity-mcp"
}
}
版本更新
UPM 通过 Git URL 安装后,会在宿主项目的 packages-lock.json 中锁定当前 commit hash。后续不会自动更新。
如需锁定特定版本,可在 URL 末尾追加 Git Tag:
https://github.com/yangfch3/Unity-MCP-Server.git#v0.1.0
对应 Packages/manifest.json 配置:
{
"dependencies": {
"com.yangfch3.unity-mcp": "https://github.com/yangfch3/Unity-MCP-Server.git#v0.1.0"
}
}
不带 Tag 则跟踪默认分支最新 commit:
{
"dependencies": {
"com.yangfch3.unity-mcp": "https://github.com/yangfch3/Unity-MCP-Server.git"
}
}
更新版本时,修改 manifest.json 中的 #tag 后缀为新版本号,或在 UPM GUI 重新 Add package from git URL 输入新 Tag 的 URL 即可。
使用
启动服务
- Unity Editor → Window → MCP Server
- 设置端口(默认 8090),点击 Start
- 复制面板中的配置 JSON
配置 Agent
将以下内容添加到 Agent 的 MCP 配置文件(如 mcp.json):
{
"mcpServers": {
"unity-mcp": {
"url": "http://localhost:8090/"
}
}
}
扩展:添加自定义工具
实现 IMcpTool 接口,放在任意 Editor 程序集中,服务启动时会自动发现并注册:
using System.Collections.Generic;
using System.Threading.Tasks;
using UnityMcp.Editor;
public class MyCustomTool : IMcpTool
{
public string Name => "my_custom_tool";
public string Category => "custom";
public string Description => "我的自定义工具";
public string InputSchema => "{\"type\":\"object\",\"properties\":{}}";
public Task<ToolResult> Execute(Dictionary<string, object> parameters)
{
return Task.FromResult(ToolResult.Success("Hello from custom tool!"));
}
}
要求
- Unity 2022.3+
- 仅 Editor 环境,不影响运行时构建
参与贡献
欢迎参与本项目的开发,详见 CONTRIBUTING.md。
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.