home / mcp / creative ideation mcp server

Creative Ideation MCP Server

Provides context-specific ideation categories and diverse options via the Gemini API for creative thinking.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "creating-cat-creative-ideation-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@creating-cat/creative-ideation-mcp"
      ],
      "env": {
        "GEMINI_MODEL": "gemini-2.5-flash-lite-preview-06-17",
        "GEMINI_API_KEY": "YOUR_GEMINI_API_KEY"
      }
    }
  }
}

You can generate context-specific creative ideas by running the Creative Ideation MCP Server. It uses the Google Gemini API to produce categorized ideas and diverse options in Japanese, with built-in error handling and rate limiting to ensure smooth interaction during idea generation.

How to use

Start the server in your MCP client to generate creative ideas. You provide an expert role, a target subject, and the desired number of categories and options. The server will return multiple categories with a set of options for each, and you can enable random sampling to mix in less conventional choices for more surprising results.

How to install

Prerequisites: Node.js 18 or newer. A Google Gemini API key is required to access the Gemini models.

1) Ensure Node.js is installed on your system. 2) Obtain a Gemini API key from Google Gemini and set it in your environment when you run the MCP server.

Install and run steps using the exact MCP command shown in the configuration example.

Configuration and startup details

The MCP server is configured to run via a package runner. Use the following standard configuration to start the server. This configuration runs the MCP package with npm/yarn tooling as shown.

{
  "mcpServers": {
    "ideation_mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@creating-cat/creative-ideation-mcp"
      ],
      "env": {
        "GEMINI_API_KEY": "YOUR_GEMINI_API_KEY",
        "GEMINI_MODEL": "gemini-2.5-flash-lite-preview-06-17"
      },
      "disabled": false,
      "timeout": 300
    }
  }
}

Notes on usage and security

- Non-deterministic results mean each run can produce different outputs even with the same parameters. Re-running can yield a broader set of ideas.

Example: run configuration for the MCP server

Use the following configuration to start the server in your environment. This example keeps Gemini API credentials in environment variables and uses the default Gemini model unless you customize it.

{
  "mcpServers": {
    "creative_ideation_mcp": {
      "command": "npx",
      "args": ["-y", "@creating-cat/creative-ideation-mcp"],
      "env": {
        "GEMINI_API_KEY": "YOUR_GEMINI_API_KEY",
        "GEMINI_MODEL": "gemini-2.5-flash-lite-preview-06-17"
      },
      "disabled": false,
      "timeout": 300
    }
  }
}

Troubleshooting and tips

If you encounter rate limit messages from the Gemini API, ensure you respect the 5-second pacing interval and retry automatically as configured.

Available tools

generate_categories

Generates a set of thought categories and associated options based on expert role, subject, and other parameters, returning structured JSON with category names, descriptions, and option lists.