Godot MCP
Enables AI assistants to interact with the Godot game engine by launching the editor, running projects, capturing debug output, managing scenes and nodes, and controlling project execution through a standardized interface.
README
Godot MCP Server
A Model Context Protocol (MCP) server that enables AI assistants to interact with the Godot game engine. Control the editor, run projects, manage scenes, and capture debug output through a standardized interface.
Features
- Debug Support: Capture output, connect to remote debugger, get real-time logs
- Seamless Integration between Godot Editor Write code in godot editor, run, and debug with AI assistance. No longer need of running in blind or without editors.
- Project Management: Launch editor, run projects, list and inspect Godot projects
- Scene Manipulation: Create scenes, add nodes, configure properties
- Asset Loading: Load sprites, textures, and export mesh libraries
- Cross-Platform: Windows, macOS, and Linux support
Installation
Prerequisites
- Godot Engine 4.x installed
- Node.js 16+ and npm
Install from npm (coming soon)
npm install -g godot-mcp
Install from source
git clone https://github.com/yourusername/godot-mcp.git
cd godot-mcp
npm install
npm run build
Configuration
For Claude Desktop
Add to your Claude Desktop configuration file:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
Linux: ~/.config/Claude/claude_desktop_config.json
{
"mcpServers": {
"godot": {
"command": "node",
"args": ["/path/to/godot-mcp/build/index.js"],
"env": {
"GODOT_PATH": "/path/to/your/godot/executable"
}
}
}
}
For Claude Code
Claude Code supports MCP servers direct configuration.
Edit .mcp.json:
{
"mcpServers": {
"godot": {
"command": "node",
"args": ["/path/to/godot-mcp/build/index.js"],
"env": {
"GODOT_PATH": "/path/to/your/godot/executable",
"DEBUG": "true"
}
}
}
}
Verify the configuration:
claude mcp list
For Cursor IDE
Cursor supports MCP servers through project-specific or global configuration.
Option 1: Project-Specific (Recommended for Godot projects)
Create .cursor/mcp.json in your project root:
{
"mcpServers": {
"godot": {
"command": "node",
"args": ["/path/to/godot-mcp/build/index.js"],
"env": {
"GODOT_PATH": "/path/to/your/godot/executable",
"DEBUG": "true"
}
}
}
}
Option 2: Global Configuration
Create ~/.cursor/mcp.json in your home directory with the same format.
Option 3: Via Cursor Settings UI
- Open Cursor Settings → Features → MCP
- Click "+ Add New MCP Server"
- Fill in:
- Name:
godot - Command:
node - Args:
/path/to/godot-mcp/build/index.js - Environment:
GODOT_PATH=/path/to/your/godot/executable
- Name:
Verify in the Output panel (Ctrl+Shift+U) → Select "MCP Logs"
Environment Variables
GODOT_PATH: Path to Godot executable (optional if Godot is in PATH or standard location)DEBUG: Set to"true"to enable detailed logging
Platform-Specific Godot Paths
Windows:
C:\Program Files\Godot\Godot.exe
macOS:
/Applications/Godot.app/Contents/MacOS/Godot
(do not include .app, just the executable path without ending of .app)
Linux:
/usr/bin/godot
/usr/local/bin/godot
/snap/bin/godot
Available Tools
The MCP server provides the following tools that AI assistants can use:
Project Management
launch_editor- Launch Godot editor for a specific projectrun_project- Run a Godot project and capture outputstop_project- Stop the currently running projectlist_projects- Find Godot projects in a directoryget_project_info- Get metadata about a Godot projectget_godot_version- Get the installed Godot version
Scene Management
create_scene- Create a new scene file with a specified root node typeadd_node- Add a node to an existing scene with optional propertiesload_sprite- Load a sprite texture into a Sprite2D nodesave_scene- Save changes to a scene fileexport_mesh_library- Export a scene as a MeshLibrary resource for GridMap
Debug & Output
get_debug_output- Get debug output and errors from a running projectconnect_remote_debugger- Connect to Godot editor's remote debugger (ports 6006/6007)get_remote_debug_output- Get output from the remote debugger connectiondisconnect_remote_debugger- Disconnect from the remote debuggercapture_screenshot- Capture a screenshot from the running game viewport
UID Management (Godot 4.4+)
get_uid- Get the UID for a specific fileupdate_project_uids- Resave resources to update UID references
Development
Build
npm run build
Watch Mode
npm run watch
Testing with MCP Inspector
npm run inspector
This launches the MCP inspector for interactive testing of all tools.
Note
Screenshot capture is still experimental and may not work in all environments.
Troubleshooting
Godot Not Found
If the server can't find Godot, set the GODOT_PATH environment variable in your MCP configuration:
{
"env": {
"GODOT_PATH": "/full/path/to/godot"
}
}
Scene File Not Created
If scenes aren't being created:
- Check directory permissions
- Verify the project path contains
project.godot - Enable debug mode to see detailed logs
Remote Debugger Connection Issues
- Ensure Godot editor is running with remote debugging enabled
- Default ports: 6007 (editor sync), 6006 (script debugger)
- Check firewall settings aren't blocking the connection
Related Projects
- Model Context Protocol - The protocol specification
- Godot Engine - The game engine
- Claude Desktop - AI assistant with MCP support
Acknowledgements
Inspired by Coding Solo's godot-mcp. https://github.com/Coding-Solo/godot-mcp
Support
For issues and questions:
- File an issue on GitHub
References
Configuration guides referenced from:
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。