Skip to content
MCP server

Tackle

MCP stdio broker that lets AI agents read and edit a running game editor's scenes

Runs locally
Install
Platforms
CLI, Linux, macOS
Runs
Locally
Version
0.3.0
Listing updated
today

Similar MCP servers

About Tackle

Tackle is an MCP server for game development. It runs locally on your machine and is started by your AI client.

Tackle sits between a running game editor and MCP clients. An editor-side package connects to a local bridge on the server, and the server then exposes the editor's scene and objects to agents over stdio using the Model Context Protocol. The README describes read tools in a U1 core and write tools in a U2 core.

The project is aimed at game developers who want an AI assistant to inspect or modify a live editor session rather than working from exported files. The product page describes the sessions as authenticated, idempotent and undo-aware, with Unity as the first backend and Godot planned later.

The package is published on the MCP Registry as ai.qompass/tackle and distributed as a Rust cargo package, with the binary named tackle.

Install

Run in a terminal

claude mcp add tackle -- npx -y tackle-mcp

Add to claude_desktop_config.json

{
  "mcpServers": {
    "tackle": {
      "command": "npx",
      "args": [
        "-y",
        "tackle-mcp"
      ]
    }
  }
}
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "tackle": {
      "command": "npx",
      "args": [
        "-y",
        "tackle-mcp"
      ]
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "tackle": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "tackle-mcp"
      ]
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "tackle": {
      "command": "npx",
      "args": [
        "-y",
        "tackle-mcp"
      ]
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "tackle": {
      "command": "npx",
      "args": [
        "-y",
        "tackle-mcp"
      ]
    }
  }
}

Using several servers? Build one config file with the MCP config generator.

What it does

Live editor bridge
Connects a running game editor to the server through a local websocket bridge.
Read tools
The U1 core exposes the editor's scene and objects to MCP clients for inspection.
Write tools
The U2 core adds tools that let agents modify editor state.
Undo-aware sessions
The product page says editor sessions are undo-aware and idempotent.
Stdio transport
Runs as a local MCP server over standard input and output.

Good for

  • Letting an AI assistant inspect the objects in an open Unity scene
  • Automating repetitive scene edits through an MCP client
  • Prototyping agent-driven level changes while keeping editor undo history

Tackle FAQ

How does the server connect to the game editor?

The editor-side package connects to the server's local bridge at ws://127.0.0.1:9770/editor.

Which game engines are supported?

The product page lists Unity as the first backend, with Godot planned for later.

How is it installed?

The MCP Registry entry lists it as a cargo package named tackle-mcp, version 0.3.2, using stdio transport.

How do I add Tackle to Claude Code?

Run this command in your terminal: claude mcp add tackle -- npx -y tackle-mcp

Where this comes from

  • Description: written from the project's README, the project's website and its MCP Registry entry.
  • Install steps: from the official MCP Registry entry.

How AIPoweredBot checks servers