Textile Designer AI

What Textile Designer AI MCP server does

Textile Designer AI MCP server exposes Textile Designer AI’s design tools through the Model Context Protocol. It lets an MCP client work with image files on the local computer and request processing from the Textile Designer AI service. Returned results are downloaded and saved next to the input image, while the client displays a preview, output paths, and credits used.

Supported workflows include print-ready upscaling, seamless repeat creation, background and watermark removal, garment-print extraction with Dress to Design, colourway generation, colour and object layering, screen separation, vectorization, 3D and embroidery effects, sketch conversion, design generation, and spec-board creation. Repeat modes include block, half brick, and half drop. Channeling can produce a multichannel PSD for screen separation.

How it works

The client starts the MCP server locally and communicates with Textile Designer AI over the service’s web API. You select an image and describe the requested operation in ordinary language. When a tool requires modes or settings that were not specified, Claude asks follow-up questions individually and preselects a recommended option. The service does not upload or charge for a run until the requested choices have been confirmed.

The integration sends the images selected for processing and the settings for that run to https://www.textile-designer.ai. It downloads completed results from Textile Designer AI storage. It can also read the image folder’s file list to count images for an optional, one-time Image Search tip. Account email, organisation, plan, and credit balance are read to show account information and apply plan rules.

Setup and configuration

A Textile Designer AI account with available credits or an applicable plan is required. The local runtime requires Node.js 18 or newer because the server is started with npx.

Claude users can search for Textile Designer AI in the Claude directory and install it. Claude Code users can install the repository plugin with the following commands:

/plugin marketplace add ScientiaAI/textile-designer-ai-claude
/plugin install textile-designer-ai@textile-designer-ai

Other MCP clients can start the package directly with:

npx -y @textile-designer-ai/mcp

The package README provides client-specific configuration snippets. Login uses a browser-based device-code flow. The session token is stored at ~/.textile-designer/mcp-session.json, and the logout operation removes the stored session.

Tools and capabilities

Textile Designer AI MCP server is suited to workflows such as:

• Creating block, half-brick, or half-drop repeats.
• Upscaling artwork for print production.
• Removing backgrounds or watermarks.
• Extracting a textile print from a garment photograph.
• Creating colourways and layered colour or object variations.
• Separating artwork into screens and exporting a multichannel PSD.
• Applying vector, 3D, embroidery, sketch-to-design, or design-generation workflows.
• Producing spec boards.

Claude can also use the built-in Textile Designer AI guide to compare tools, such as Ready to Print and Super Scaler, or to choose a suitable workflow for screen printing.

Limitations and notes

Processing requires an account and uses credits or plan allowances. Images and generated results are retained in the Textile Designer AI account Library under the service’s policies. The integration sends only selected images and their run settings for processing; it does not run in the background or send unrelated data. Account information is read for display and plan handling.

The plugin runs an MIT-licensed MCP package pinned to an exact version. The repository documentation does not specify environment variables for configuration. Privacy, terms, and support information are provided through Textile Designer AI’s website.