Unity AI Bridge
Lets AI IDEs control the Unity Editor via file-based IPC, for Unity developers
- Pricing
- Open source
- Platforms
- Windows, macOS, vscode
- Version
- 1.0.0
- Listing updated
- today
Similar MCP servers
About Unity AI Bridge
Unity AI Bridge is an open-source MCP server for developer tools. It is available on Windows, macOS, vscode, CLI and API.
The README says Unity AI Bridge gives AI coding assistants direct access to the Unity Editor, which is otherwise invisible to them. Through a Unity package, an AI can inspect and edit scenes, GameObjects, assets and prefabs, run tests, read the console and capture profiler data, instead of only reading and writing source files.
The bridge talks to the editor through a pair of request and response files on disk rather than network sockets, so it opens no ports. The same Unity plugin is reached either through a Python command-line tool used in Skill mode or through a Python MCP server, and the README lists Claude Code, Cursor, GitHub Copilot, Windsurf and Claude Desktop as supported clients.
The project is aimed at Unity teams who want AI agents to run an edit, test and fix loop inside the editor, including play-mode checks with runtime queries and screenshots. Custom tools can be added by tagging static C# methods with an attribute, and the README reports Windows and macOS as the supported platforms.
Tools this server provides (10)
Taken from the project documentation; the running server may differ.
- scene-open
- Open a scene in the Unity Editor.
- gameobject-create
- Create a GameObject in the open scene.
- assets-find
- Search project assets.
- script-update-or-create
- Create or update a C# script.
- runtime-invoke
- Call static methods during play mode.
- runtime-query
- Read MonoBehaviour fields at runtime.
- screenshot-capture
- Capture a screenshot of the editor or game view.
- console-get-logs
- Read Unity console logs.
- tests-run
- Run Unity tests.
- profiler-snapshot
- Take a profiler snapshot.
What it does
- Broad editor tool coverage
- Exposes 65 tools across scene, GameObject, asset, prefab, profiler, light probe and test operations.
- File-based editor communication
- Uses request and response files on disk instead of open network ports or WebSocket connections.
- Play-mode testing loop
- Lets an AI enter Play Mode, invoke game methods, read runtime state and capture screenshots.
- Custom tool attribute
- Makes static C# methods callable by AI when they are tagged with a bridge tool attribute.
- Multiple IDE support
- Works with Claude Code, Cursor, GitHub Copilot, Windsurf and Claude Desktop according to the README.
- No external dependencies
- The Python server uses only the standard library and the Unity side is a self-contained package.
Good for
- Letting an AI assistant edit scenes, GameObjects and materials directly in the Unity Editor
- Running automated play-mode tests and having an AI judge the results from state and screenshots
- Profiling a Unity project and having an AI analyse snapshots or hotpaths
- Baking light probes or running Unity tests from an AI IDE
- Adding project-specific editor tools that an AI can call
Pricing
Open source
Free and open source under Apache License 2.0.
Unity AI Bridge FAQ
Does it open network ports?
No. The README says it uses file-based IPC and opens no network listeners, with the channel scoped to the user's temp directory.
Which Unity versions are supported?
The README lists Unity 2022.3 LTS and later, including Unity 6, with Built-in, URP and HDRP render pipelines.
Which operating systems are supported?
The README lists Windows and macOS as platforms.
What license does it use?
The repository is licensed under the Apache License 2.0, and the README notes it is derived from Unity-MCP under the same license.
Is Unity AI Bridge free?
Yes. Unity AI Bridge is open source, so you can use and self-host it for free.
What are the best alternatives to Unity AI Bridge?
The closest alternatives to Unity AI Bridge are unity-mcp, Unity MCP, Funplay Unity MCP, AI Game Developer: MCP for Unity, Unreal and Godot and KitWright Unity MCP.
Where this comes from
- Description: written from the project's README and its MCP Registry entry.
