Official Announcement | OBR Group Launches the First Open-Source Version of ArtForge AI MCP

OBR Group — Open Source Release v1.0.0
OBR Group officially announces the release of the first open-source version of ArtForge AI MCP, a specialized drawing and design component developed to work with artificial intelligence agents through the Model Context Protocol (MCP).
This release follows multiple stages of development and experimentation, with the aim of making the project available to developers and enthusiasts to explore its capabilities, contribute improvements, and participate in its continued development.

First: What Is ArtForge AI MCP?

It is an MCP server that connects artificial intelligence models to a 2D drawing engine, enabling users to create designs and modify their elements through specialized programming tools instead of relying entirely on conventional AI image generation.
The project supports both Arabic and English, with particular attention to Arabic text rendering, character shaping, and right-to-left (RTL) writing direction.

Second: Skills and Capabilities

The component provides skills and tools for creating advertising posters, cards, content covers, social media posts, text-based designs, and graphical scenes.
It also includes bundled Arabic fonts, such as Amiri, Tajawal, and Almarai, along with tools for controlling colors, gradients, layers, shapes, and text.
Additionally, it supports the use of design references and color systems to improve design organization and guide drawing operations.

Third: The Drawing Engine

The project is powered by @napi-rs/canvas, using Skia and HarfBuzz technologies for graphics and text processing.
These technologies provide vector drawing capabilities, anti-aliasing, Arabic character shaping, and control over dimensions and visual formatting.
The published version supports exporting designs in PNG, JPEG, and WebP formats.

Fourth: Programming Tools

The component includes several specialized tools, including:
artforge_create_design: Creates a design from a text description, including preview generation and output inspection.
artforge_plan_design: Plans the design and defines its elements.
artforge_execute_design: Executes the drawing plan.
artforge_tool_op: Performs drawing and editing operations, such as adding text, shapes, and gradients.
artforge_export: Exports the completed design as an image.
artforge_capabilities: Displays the component's available capabilities and skills.

Fifth: Connecting with ChatGPT and Claude

Users can explore connecting the component to ChatGPT, Claude, or any client compatible with the Model Context Protocol (MCP), subject to each platform's connection and authentication requirements.
A dedicated developer page provides project information and connection instructions.
Project and Connection Page:
https://obr.obr432.workers.dev/ai-plugins/

Sixth: Open-Source Project

The first version is now available on GitHub under the Apache License 2.0, allowing developers to explore, modify, and contribute to the source code in accordance with the license terms.
GitHub Repository — Open-Source Release Branch:
https://github.com/duhbobaldo-oss/artforge-mcp-trial/tree/chore/open-source-readiness
We welcome developers to test the project, report bugs, suggest improvements, and submit code contributions.

Release Notes

This is an initial open-source release intended for experimentation and development, not the complete version used within OBR Group's internal development environment.
Output quality may vary depending on the AI model used and the instructions provided. Some advanced capabilities are also unavailable in this release.
Our vision at OBR Group is to develop specialized tools that help artificial intelligence models move beyond generating answers toward performing practical tasks that can be tested, developed, and continuously improved.
Developed by OBR Group
ArtForge AI MCP — Open Source v1.0.0

Important Note: Getting the Best Results with ArtForge AI MCP
ArtForge AI MCP includes specialized design skills, a real drawing engine, color palettes, and advanced drawing tools.
For the best results, provide detailed prompts specifying colors, dimensions, fonts, spacing, drawing styles, and visual elements.
Giving the AI agent complete freedom may work occasionally, but my testing showed that precise instructions produce more reliable results.
You can also ask the agent to search the web for visual references and recreate the desired concept using ArtForge's drawing tools.
My recommendation: Treat every prompt as a detailed design brief to achieve more accurate and consistent results.