UI Color Palette MCP

Connect any AI assistant to UI Color Palette via the Model Context Protocol

The UI Color Palette MCP is a Model Context Protocol server that exposes the UI Color Palette engine to any MCP-compatible AI assistant — Claude, Cursor, Windsurf, and others.

It lets an AI generate, preview, audit, and publish color palettes without opening a design tool.

Endpoint

https://mcp-uicp.yelbolt.workers.dev/mcp

The server runs on Cloudflare Workers with durable state per session.

Tools at a glance

Palette tools — no authentication required

Tool	Description
get_palette	Generate a complete color palette from a base configuration and themes
get_color_system	Build a semantic color system from taxonomy bindings
create_color_harmony	Generate color harmonies from a base color
extract_dominant_colors	Extract dominant colors from an image URL
generate_code	Export palette tokens in CSS, SCSS, Tailwind, Swift, and more
generate_colors_from_prompt	Generate source colors from a natural language description
preview_palette	Render a palette as an inline image

Community tools — authentication required

Tool	Description
list_published_palettes	Browse publicly shared palettes
list_my_published_palettes	List your own published palettes
get_published_palette	Fetch a specific palette by ID
publish_palette	Save a palette to the database
update_published_palette	Update a published palette
share_published_palette	Make a palette public in the community
unshare_published_palette	Make a palette private
unpublish_palette	Permanently delete a published palette

Installation

Claude Desktop

Add the server to claude_desktop_config.json:

{
  "mcpServers": {
    "ui-color-palette": {
      "url": "https://mcp-uicp.yelbolt.workers.dev/mcp"
    }
  }
}

Claude Code

Add the server to ~/.claude/settings.json (user-wide) or .claude/settings.json (project-level):

{
  "mcpServers": {
    "ui-color-palette": {
      "type": "http",
      "url": "https://mcp-uicp.yelbolt.workers.dev/mcp"
    }
  }
}

Or use the CLI directly:

claude mcp add --transport http ui-color-palette https://mcp-uicp.yelbolt.workers.dev/mcp

VS Code

Add the server to .vscode/mcp.json in your project (or the equivalent user-level config):

{
  "servers": {
    "ui-color-palette": {
      "type": "http",
      "url": "https://mcp-uicp.yelbolt.workers.dev/mcp"
    }
  }
}

Cursor

Add the server to ~/.cursor/mcp.json (user-wide) or .cursor/mcp.json (project-level):

{
  "mcpServers": {
    "ui-color-palette": {
      "url": "https://mcp-uicp.yelbolt.workers.dev/mcp"
    }
  }
}

Windsurf

Add the server to ~/.codeium/windsurf/mcp_config.json:

{
  "mcpServers": {
    "ui-color-palette": {
      "serverUrl": "https://mcp-uicp.yelbolt.workers.dev/mcp"
    }
  }
}

OpenAI Codex

Add the server to ~/.codex/config.json:

{
  "mcpServers": {
    "ui-color-palette": {
      "url": "https://mcp-uicp.yelbolt.workers.dev/mcp"
    }
  }
}

Gemini CLI

Add the server to ~/.gemini/settings.json:

{
  "mcpServers": {
    "ui-color-palette": {
      "httpUrl": "https://mcp-uicp.yelbolt.workers.dev/mcp"
    }
  }
}

---

Quick start

1. Add the server to your client — see Installation above.

2. Authenticate (optional)

Community tools require a UI Color Palette account. The server implements OAuth 2.0 — your MCP client will handle the flow automatically when you first call an authenticated tool. See Authentication for details.

3. Generate a palette

Ask your AI assistant:

"Generate a Tailwind palette from a deep forest green (#2D6A4F), light and dark themes, and show me a preview."

The assistant will call get_palette, then preview_palette to render the result inline.