SudoMock Now Works with Claude, Cursor, ChatGPT, and Any MCP-Compatible AI Tool

SudoMock's MCP server is live. You can now render product mockups directly from any AI assistant that supports the Model Context Protocol. Claude, Cursor, VS Code, ChatGPT, Windsurf, and any future MCP-compatible tool can connect.

If you've been copy-pasting curl commands or writing fetch wrappers to generate mockups, those days are over. This post covers what MCP is, how to connect SudoMock to your AI tools, what you can build with it, and why we think this changes how developers interact with rendering APIs.

#What is MCP?

MCP stands for Model Context Protocol. Anthropic created it as a standard way for AI assistants to call external tools. Think of it as a USB port for AI: any tool that speaks MCP can plug into any AI assistant that supports it.

Before MCP, connecting an AI assistant to an external API meant writing custom integration code. You'd need to describe the API endpoints to the model, handle authentication, parse responses, and deal with errors. Every new API meant another custom integration.

MCP changes this. The protocol defines a standard interface: the server describes its tools (with parameters, types, and descriptions), and the AI client discovers them automatically. The client knows what tools are available, what parameters they accept, and how to call them. No glue code needed.

For mockup rendering, this is a big deal. A typical render workflow involves 3-4 API calls: list templates, get Smart Object UUIDs, send the render request, retrieve the output URL. With MCP, your AI assistant handles the entire chain from a single natural language instruction.

SudoMock is one of the first mockup APIs to support MCP natively. Claude, Cursor, and the other supported clients can call our render API, manage your templates, and check your account balance without you writing a single line of integration code.

#What you can do with SudoMock MCP

The MCP server exposes 9 tools organized into four categories:

Rendering

• render_mockup (1 credit) places your artwork onto a PSD template and returns a CDN URL
• ai_render (5 credits) uses AI to detect printable surfaces on any product photo. No PSD template needed. Just point it at a product photo and your design. Learn more about SudoAI Render.

Template Management

• list_mockups lists your uploaded templates with UUIDs and thumbnails
• get_mockup_details returns Smart Object UUIDs, layer names, dimensions, and blend modes
• upload_psd uploads a new PSD/PSB file as a mockup template
• update_mockup renames a template
• delete_mockup permanently removes a template and its associated files

Account

• get_account shows your plan, credit balance, usage stats, and billing period

Studio

• create_studio_session launches an interactive mockup editor for customer-facing customization

All management operations (list, get, update, delete) are free. You only spend credits on actual renders.

Here's what a typical conversation looks like in practice:

"Show me my mockup templates"

The AI calls list_mockups and returns your template library with names, thumbnails, and UUIDs.

"Render my logo on the t-shirt mockup in WebP format"

The AI calls render_mockup with your template UUID, your logo URL, and image_format: "webp". You get a CDN URL back in under a second.

"Now do the same for all my hoodie mockups at 2x size"

The AI iterates through your templates, filters hoodies by name, and fires parallel render calls. Five renders, five CDN URLs, done.

"How many credits do I have left?"

The AI calls get_account and tells you your remaining balance, current plan, and when credits reset.

#Supported AI clients

SudoMock MCP works with six major AI coding assistants:

• Claude Code and Claude Desktop by Anthropic
• Cursor, the AI-first code editor
• VS Code with GitHub Copilot's MCP support
• ChatGPT by OpenAI
• Windsurf by Codeium

Each client connects via one of two methods:

Remote (OAuth) connects directly to mcp.sudomock.com. Your AI client opens a browser window for authentication. You log in with your SudoMock account, authorize the connection, and you're done. No API key to copy or manage. Tokens refresh automatically in the background.

Local (npx) runs a local MCP process using your API key. This method works offline, in CI/CD pipelines, and in environments without a browser. Install the @sudomock/mcp package from npm and provide your key as an environment variable.

Both methods give access to the same 9 tools. Pick whichever fits your workflow.

#Connect in 60 seconds

Pick your client and follow the instructions. Each one takes under a minute.

Claude Code

The fastest setup. One command in your terminal:

claude mcp add --transport http sudomock https://mcp.sudomock.com

# Or local with API key
claude mcp add sudomock -e SUDOMOCK_API_KEY=sm_your_key -- npx -y @sudomock/mcp

Cursor

Add to .cursor/mcp.json in your project root:

{
  "mcpServers": {
    "sudomock": {
      "type": "http",
      "url": "https://mcp.sudomock.com"
    }
  }
}

Or use the one-click install button on our MCP documentation page. It opens Cursor and adds the server automatically.

VS Code

Add to .vscode/mcp.json in your workspace:

{
  "servers": {
    "sudomock": {
      "type": "http",
      "url": "https://mcp.sudomock.com"
    }
  }
}

Claude Desktop and ChatGPT

Add the server URL to your client's settings. Both handle OAuth automatically when you connect to https://mcp.sudomock.com. Detailed config examples are in our MCP docs.

That's it. No SDK to install, no webhook to configure, no callback URLs to register.

#Real-world workflow: batch rendering

Here's a scenario we see constantly. You have 20 t-shirt mockup templates and a new logo. You want all 20 rendered in WebP at 1920px wide.

Without MCP, you'd write a script. Loop through template UUIDs, construct JSON request bodies, send POST requests, handle rate limits, collect output URLs. Maybe 30-40 lines of Python or JavaScript, plus error handling and retry logic. You'd also need to look up Smart Object UUIDs for each template first.

With MCP, you type one sentence:

"Render my new logo on all t-shirt mockups. WebP format, 1920px. Give me the URLs when done."

Claude calls list_mockups, filters by name, looks up Smart Object UUIDs with get_mockup_details, then fires 20 parallel render_mockup calls (respecting your plan's concurrency limit). Each render completes in under a second. You get 20 CDN URLs in a single response.

Total cost? 20 credits. That's $0.04 on the Scale tier.

This pattern scales to any batch operation: updating product images across your Shopify store, generating seasonal design variations, A/B testing layout options, or preparing listings for multiple marketplaces at once.

And it works for SudoAI Render too. Point it at 10 product photos and a design, and the AI handles segmentation, perspective correction, and compositing for each one. 50 credits for 10 AI renders that would take hours manually.

#Two authentication methods

OAuth 2.1 (recommended for interactive use)

The remote server at mcp.sudomock.com handles OAuth 2.1 automatically. When your AI client connects for the first time, it opens a browser window. You log in with your SudoMock account and authorize the connection. Tokens refresh in the background. You never need to manage keys or credentials.

API Key (recommended for automation)

For CI/CD pipelines, background jobs, or environments without a browser, use an API key. Create one at sudomock.com/dashboard/api-keys and pass it as an environment variable:

export SUDOMOCK_API_KEY="sm_your_key_here"

# Connect via npx
claude mcp add sudomock -e SUDOMOCK_API_KEY=$SUDOMOCK_API_KEY -- npx -y @sudomock/mcp

Your key starts with sm_ and gives full access to your account's tools and credits. Store it in environment variables or a secrets manager. Never commit it to source control.

#Available everywhere MCP runs

We've registered SudoMock on the official MCP Registry, Smithery, mcp.so, and other MCP directories. If your AI tool supports browsing MCP servers, you can find us there.

We also publish SDKs for traditional API integration:

• Python SDK on PyPI (pip install sudomock)
• Node.js SDK on npm (npm install sudomock)

Both are typed, support async operations, and include error handling. Use them when you need programmatic control beyond what natural language provides.

#Get started

SudoMock's MCP server is free to try. Sign up and get 500 credits to test your workflow. No credit card required. Paid plans start at $17.49/month with 5,000 credits.

Here's the fastest path:

1. Pick your AI client from the list above
2. Add the MCP server (one command or one config line)
3. Say "show me my mockup templates" to verify the connection
4. Follow our Quick Start guide to upload your first PSD and render a mockup

Full tool documentation, per-client setup guides, and example prompts are in our MCP documentation.