`Godot MCP` is an AI-powered game development assistant **for the Godot Editor**. Connect **Claude**, **Cursor**, **Copilot**, or any MCP-aware agent to Godot and let it inspect and drive your project — create nodes, edit scenes, manage resources and scripts, capture screenshots, and more.
Godot-MCP is the Godot counterpart of [Unity-MCP](https://github.com/IvanMurzak/Unity-MCP): a C# **editor addon** that exposes Godot Editor operations as **AI Tools** and connects them to an MCP server through the same hosted cloud backend ([ai-game.dev](https://ai-game.dev)) that powers Unity-MCP — or your own self-hosted server. The MCP / reflection stack is **not forked**: it is shared with Unity-MCP and consumed from [nuget.org](https://www.nuget.org/) as `PackageReference`s.
> **[💬 Join our Discord Server](https://discord.gg/cfbdMZX99G)** — Ask questions, showcase your work, and connect with other developers!
## 
- ✔️ **AI agents** — Use the best agents from **Anthropic**, **OpenAI**, **Google**, or any other provider with no vendor lock-in
- ✔️ **23 built-in Tools** — [MCP Tools](#tools-reference) across 9 families for operating the Godot Editor, led by [RQL](docs/rql.md): one query language for the whole editor
- ✔️ **C# & GDScript** — Read, create, and update both `.cs` and `.gd` scripts, and attach them to nodes
- ✔️ **Scene & Node control** — Query and edit the scene tree, Godot properties (GDScript variables included), `.tscn` scenes and `.tres`/`.res` resources with RQL — bulk, all-or-nothing, undoable
- ✔️ **Visual feedback** — Capture viewport, camera, and isolated-node screenshots the LLM can inspect
- ✔️ **Call anything** — `call CallMethod("name", [args])` reaches any Godot method (GDScript ones too); in a running game, find and call C# methods via [ReflectorNet](https://www.nuget.org/packages/com.IvanMurzak.ReflectorNet)
- ✔️ **Cloud or self-hosted** — Connect to `ai-game.dev` out of the box, or point at your own server
- ✔️ **Natural conversation** — Chat with AI like you would with a human

# Quick Start
Get up and running from a terminal using the [`godot-cli`](https://www.npmjs.com/package/godot-cli) (the Godot analog of `unity-mcp-cli`) — no manual file copying or csproj editing required:
```bash
# 1. Install godot-cli
npm install -g godot-cli
# 2. (Optional) Scaffold a fresh Godot C# project — skip if you already have one
godot-cli create-project --dotnet ./MyGodotProject
# 3. Install the godot_mcp addon: downloads addons/godot_mcp/ from the matching
# GitHub release, adds the required NuGet packages + the extension-catalog
# EmbeddedResource to your .csproj, and enables the plugin in project.godot —
# all idempotently
godot-cli install-plugin ./MyGodotProject
# 4. Sign in to the ai-game.dev cloud — OAuth 2.1 device login (opens a browser,
# saves a machine-wide credential the editor plugin auto-adopts; no token to copy)
godot-cli login
# 5. Pick an AI agent (Claude Code, Cursor, Copilot, …) and write its MCP config —
# pinned to THIS project's cloud route by default (pass --no-pin for the bare URL)
godot-cli setup-mcp claude-code ./MyGodotProject
# 6. Open the Godot editor — builds the C# assembly first (so the addon loads on
# the very first open) then auto-connects with the right GODOT_MCP_* env vars
godot-cli open ./MyGodotProject
# 7. Wait until the plugin answers the readiness probe
godot-cli wait-for-ready ./MyGodotProject
```
That's it. Ask your AI *"Create 3 cubes in a circle with radius 2"* and watch it happen. ✨
> **Offline / dev install:** `install-plugin --source