Cheat Engine MCP Server

Warning

This project is not feature complete and is under heavy active development.

Things might break while features are added.

A Model Context Protocol (MCP) server plugin for Cheat Engine that provides access to Cheat Engine functionality.

Architecture

This project exposes Cheat Engine functionality through a REST API server built with ASP.NET Core.

• REST API Server: Runs on http://localhost:6300 with OpenAPI documentation at /scalar/v1
• 17+ API Endpoints: Lua execution, process management, memory operations, AOB scanning, disassembly, and more
• Single DLL Plugin: All dependencies embedded - just drop into Cheat Engine plugins folder

Related Projects

• Python MCP Client: hedgehogform/ce-mcp-client - Wraps the REST API with MCP SSE server for AI clients like Claude Desktop

Requirements

• Cheat Engine 7.6.2+ (minimum version with .NET Core plugin support)
• .NET 9.0 SDK
• Windows OS (I don't have a mac but if you can run it then open an issue and let me know.)

Important

Cheat Engine 7.6.2 or newer is required. Older versions do not support .NET Core plugins.

Installation

1. Build the project: dotnet build
2. Copy ce-mcp.dll from bin/ to your Cheat Engine plugins directory
3. Enable the plugin in Cheat Engine

Development

Initial Setup

First, initialize the git submodule (CESDK):

git submodule update --init --recursive

If you cloned the repo without submodules, this command will download the required CESDK dependency.

Building

# Build the C# plugin
dotnet build

# Build in Release mode
dotnet build -c Release

Note: If you encounter a FodyCommon.dll access denied error during restore/build, close your IDE and restart it to release the file lock.

Python MCP Client

# Navigate to mcp-client directory
cd mcp-client

# Install dependencies
uv sync

# Open cheat_engine_mcp_server.py in your AI software.

Testing

1. Build the plugin and copy to Cheat Engine plugins directory
2. Start Cheat Engine and enable the plugin
3. Use "MCP" menu to start the server
4. Access the REST API at http://localhost:6300 or view documentation at http://localhost:6300/scalar/v1
5. For AI integration, use ce-mcp-client to connect Claude Desktop or other MCP clients

Available API Endpoints

For complete API documentation with interactive examples, visit http://localhost:6300/scalar/v1 after starting the server.