Happy slice of mcp pizzaMCP.pizza

gimp-mcp

MCP.Pizza Chef: libreearth

Two pieces make this work: a plugin you drop into GIMP 3, and a small program your chat app launches. Together they offer four commands — list open images, read an image's size and layers, apply a Gaussian blur, and a catch-all that calls any GIMP function by name. The project page promises background removal, inpainting and style transfer, but none of that is in the code. Nothing is charged and no account is needed. The repository was published in March 2025 and has not changed since.

Unmaintained · No commits in 18 months.
Images/Design

Use This MCP server To

List every image I currently have open in GIMP Check the size and layer count of an open image Run a named GIMP command without hunting through menus Automate a repetitive GIMP step by asking for it

README

GIMP-MCP: Integrating GIMP with AI through Model Context Protocol

GIMP-MCP is an initiative to integrate the Model Context Protocol (MCP) into the GNU Image Manipulation Program (GIMP), enabling seamless interaction between GIMP and AI models. This integration allows users to harness advanced AI capabilities directly within GIMP, enhancing image editing workflows with intelligent automation and context-aware operations.

Table of Contents

Introduction

The Model Context Protocol (MCP) is an open standard developed to facilitate seamless integration between AI systems and external applications. By implementing MCP within GIMP, users can leverage AI models for tasks such as object recognition, style transfer, and automated enhancements, thereby extending GIMP's native functionalities.

Features

  • AI Integration: Connect GIMP with state-of-the-art AI models for enhanced image processing capabilities.
  • Automated Workflows: Utilize AI to perform repetitive tasks, allowing for more efficient editing processes.
  • Context-Aware Operations: AI models can interpret and manipulate images based on contextual understanding, leading to more intelligent edits.

Installation

Prerequisites

Before integrating MCP with GIMP, ensure you have the following:

  • GIMP 2.10 or later installed on your system.
  • Python 3.x installed.
  • Basic knowledge of Python scripting and GIMP's plugin architecture.

gimp-mcp FAQ

Is this project still being worked on?
No. Everything in the repository was pushed on a single day in March 2025 and nothing has changed since, so treat it as an experiment rather than a maintained tool.
Can I use this to remove a background or repaint part of a photo?
No. The project page advertises background removal, inpainting and style transfer under a Filters > AI Tools menu, but none of that exists in the code — the only editing command included is a Gaussian blur.
Do I need a paid account or a key?
No. No credential is read anywhere in the code and nothing is sent to an outside service; everything runs against GIMP on your own machine.
Which version of GIMP does it need?
GIMP 3. The written instructions say 2.10, but the plugin asks the system for GIMP 3.0 on its first line, so it will not load in 2.10.
How hard is the setup?
Harder than the instructions suggest. They tell you to download a different project entirely and run a file that is not in it. In practice you copy the plugin into GIMP's plug-in folder, make it executable, start it from Filters > Development, then run the second file as your server.
Do the ready-made commands actually work?
Expect trouble. The three shortcuts send names with an extra 'Gimp.' in front that the plugin cannot look up, and the blur command also asks for a GIMP feature that was removed in version 3, so it comes back as an error.
Is there anything to be careful about?
Yes. While it is running, the plugin accepts instructions on your machine with no password and no limit on what it will do, including loading and saving files, so any other program on the same computer can drive GIMP through it.
Can I use it at work?
Check with whoever handles licensing first. The read-me claims MIT, but the license file in the repository is GPL-3, which carries much stronger sharing obligations.