DesignStudio MCP

An MCP server implementing an AI creative-director persona: produces full production-ready design concepts, evaluates brand names, runs competitor visual analysis, analyzes uploaded images via a Groq vision model, and generates platform-specific AI image prompts (Flux, GPT Image, Midjourney, Ideogram, Stable Diffusion).

Setup

npm install
npm run build

Set GROQ_API_KEY in your environment (see .env.example). Get a key at https://console.groq.com. The server works without a key too — every tool falls back to a deterministic, decisive local recommendation engine instead of failing or returning a generic refusal.

Connect to an MCP client (e.g. Claude Desktop)

Add to your client's MCP config (e.g. claude_desktop_config.json):

{
  "mcpServers": {
    "designstudio": {
      "command": "node",
      "args": ["/absolute/path/to/designstudio-mcp/build/index.js"],
      "env": {
        "GROQ_API_KEY": "your_groq_api_key_here"
      }
    }
  }
}

Tools exposed

• create_design_concept — full creative-director output (the main tool)
• evaluate_brand_names — scores/ranks candidate brand names
• competitor_visual_analysis — visual differentiation strategy vs named competitors
• analyze_uploaded_image — Groq vision analysis of an uploaded logo/asset/screenshot
• generate_ai_image_prompts — platform-specific AI image prompts from a finalized concept

Notes on scope

This is a real MCP server (stdio transport, validated against the MCP protocol). It does NOT render actual images — "AI image generation" tools here return optimized prompts for you to paste into Flux/Midjourney/etc. Wiring direct image generation would require separate API keys per image provider and is a natural next extension (see groqClient.ts for the pattern to follow).