MCP Catalogs
Homefunplay-unity-mcp screenshot

funplay-unity-mcp

by FunplayAI·57·Score 46

Advanced MCP server for Unity Editor with 91 built-in tools for automation, scene creation, and AI development.

developer-toolsai-llmproductivity
5
Forks
0
Open issues
this month
Last commit
2d ago
Indexed

Overview

Funplay MCP for Unity is a comprehensive MCP server that enables AI assistants like Claude Code, Cursor, and VS Code Copilot to operate directly within Unity projects. It provides an execute_code functionality for in-memory C# execution, 91 built-in tools across various categories, and supports play mode automation with input simulation and screenshots. The server features a clean architecture with both core and full tool profiles, allowing users to select the appropriate level of functionality for their needs.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

you:Automating Unity scene creation and script generation through natural language prompts
you:Testing and validation of gameplay mechanics through input simulation and screenshot capture
you:Performance analysis and debugging of Unity projects with automated inspection tools
you:What is the difference between 'core' and 'full' tool profiles?
you:Does this package work with any AI client that supports MCP?

When to choose this

Choose this when working with Unity projects and need AI agents to directly interact with the editor through comprehensive tooling for scene creation, script generation, and automation.

When NOT to choose this

Avoid if you need runtime game functionality (this is Editor-only) or if you prefer a Python-based MCP server architecture rather than a Unity package.

Tools this server exposes

12 tools extracted from the README
  • execute_code

    Execute C# code in-memory within Unity Editor

  • get_scene_info

    Retrieve information about the currently open Unity scene

  • unity://project/context

    Access the current Unity project context including scene, selection, and compilation state

  • play_mode_control

    Control Unity play mode (enter, exit, pause)

  • simulate_input

    Simulate keyboard and mouse input in Unity play mode

  • capture_screenshot

    Take screenshots of the game view or scene view

  • get_performance_data

    Retrieve Unity performance metrics and frame data

  • get_console_output

    Read Unity console logs and error messages

  • find_objects_by_type

    Locate Unity objects by type or component

  • edit_component

    Modify components on GameObjects using SerializedObject

  • check_compilation

    Check if Unity scripts compile successfully

  • execute_menu_item

    Execute Unity menu items programmatically

Comparable tools

unity-mcpplaywright-mcpshell-mcp

Installation

  1. Install via UPM (Git URL): In Unity, go to **Window → Package Manager → + → Add package from git URL** and enter https://github.com/FunplayAI/funplay-unity-mcp.git
  2. Start the MCP Server: Use **Menu: Funplay → MCP Server** to start the server on http://127.0.0.1:8765/
  3. Configure Your AI Client: Use the built-in **One-Click MCP Configuration** or manually add:
{
  "mcpServers": {
    "funplay": {
      "type": "http",
      "url": "http://127.0.0.1:8765/"
    }
  }
}

FAQ

What is the difference between 'core' and 'full' tool profiles?
The 'core' profile exposes 29 high-signal tools focused on essential functionality, while 'full' exposes all 91 tools. Users can switch between profiles in the MCP Server window.
Does this package work with any AI client that supports MCP?
Yes, it's vendor agnostic and works with Claude Code, Cursor, Windsurf, Codex, VS Code Copilot, and any other MCP-compatible client.

Compare funplay-unity-mcp with

GitHub →

Last updated · Auto-generated from public README + GitHub signals.