Updated:
September 12, 2025
Sparks

Turning prompts into brand-perfect pictures

What if you could ask your brand assistant to make an image Instagram-ready—and it just knew what to do? That was the spark behind our first-ever Model Context Protocol (MCP) server, built during Frontify’s annual innovation week. In just one week, we wrapped our Media Processing API into a lightweight MCP server. The goal? Let any AI agent—like ChatGPT, Claude, or within Cursor—access Frontify’s asset transformation powers using nothing but natural language. Resize, crop, format, filter—now accessible as AI-native tools.
On this page

What’s an MCP Server, exactly?

The Model Context Protocol1 is an emerging standard that allows AI models to interact with external tools in a structured way. Think of it as a translator that turns a human prompt into a precise API call. By building an MCP server for Frontify’s Media Delivery product2, we created a bridge between our asset transformation capabilities and AI systems. The result? AI can now understand how to manipulate your brand assets through simple, intuitive prompts—no developer required.

Seeing it in action

  • With Claude.ai: We prompted it to make an image optimized for an Instagram post. Claude used our MCP server to generate the exact transformed CDN link—and explained every change it made.
  • On Frontify: We embedded the MCP tools within the Media Transformer app, letting users transform images directly from their asset library with simple prompts.
  • Using Cursor.ai: A developer transformed an entire image gallery through chat—resizing, cropping, changing backgrounds—without touching a single line of manual code.

Why it matters

When you remove the distance between a bright idea and a brand-perfect asset, everything accelerates. Designers iterate without leaving the conversation, marketers launch variants without tapping a production queue, and engineers stop writing one-off image scripts just to keep up. The MCP server makes every prompt a direct line to brand-safe visuals, so teams trade friction for flow and speed for scale—while the brand stays unmistakably yours.

We’re betting big on the MCP standard. This early experiment proved that bringing AI and brand assets together isn’t just possible—it’s powerful. With just one protocol, we unlocked instant, two-way collaboration between AI tools and your brand visuals.

Get your hands on it

Being a Frontify customer you can use the hosted version of the MCP server. To add it to your Cursor.AI IDE, you will just need to add following code to your MCP settings:

 "mcpServers": {
    "frontify-media-delivery": {
      "url": "https://mcp.frontify.dev/mcp"
    },
}

You can also try it out via App: within your Frontify environment head to the Marketplace and install the “Media Transformer” app.

Media Transformer app on Frontify Marketplace

When every prompt can produce a brand-perfect picture, imagination scales with your ambition.

No items found.
decorative
Daniel Demel
Principal Product Innovation

Brand news straight to your inbox:

By subscribing you agree to our TOS and Privacy Policy