1. Create Your Account

Start by creating a free account to access all features:

• Click Register in the top navigation
• Choose email registration or OAuth (GitHub/Google)
• Verify your email if using email registration
• Complete your profile setup

2. Browse Existing Servers

Explore the marketplace to discover community-built MCP servers:

• Use the search bar to find servers by name or functionality
• Filter by categories, tags, or popularity
• View detailed server information, installation instructions, and reviews
• Copy installation commands with one click

3. Try the AI Generator

Generate custom MCP servers instantly using AI:

• Navigate to the Generate page
• Describe your desired server functionality in natural language
• Choose your preferred AI model (Claude, GPT-4, Gemini, or Grok)
• Review and download the generated code

Step 1: Navigate to Generator

Click on Generate in the navigation menu to access the AI generator.

Step 2: Configure API Key

Before generating, ensure you have configured at least one AI provider API key in Settings.

Step 3: Describe Your Server

Write a clear, detailed description of what you want your MCP server to do. Be specific about:

• The main functionality or purpose
• Data sources or APIs to integrate
• Expected inputs and outputs
• Any specific requirements or constraints

Step 4: Select AI Model

Choose your preferred AI model based on your needs:

Step 5: Generate & Review

Click Generate Server and wait for the AI to create your code. The generation typically takes 10-30 seconds.

Review the generated code carefully:

• Check that all required dependencies are included
• Verify the code matches your requirements
• Review any configuration options or environment variables
• Test the code locally before publishing

Step 6: Download or Publish

You can either:

• Download the code to use privately
• Publish to the marketplace to share with the community
• Regenerate if the output doesn't meet your needs

Before Publishing

Ensure your server meets these requirements:

• Code is well-tested and functional
• Clear documentation and usage examples
• Proper error handling and validation
• Security best practices implemented
• Dependencies clearly listed
• Open-source license included (MIT, Apache 2.0, etc.)

Publishing Steps

# Clone the repository
git clone https://github.com/username/mcp-server-name
cd mcp-server-name

# Install dependencies
npm install

# Configure environment
cp .env.example .env
# Edit .env with your settings

# Run the server
npm start

Review Process

Submissions are typically reviewed within 24-48 hours. Reviewers check for:

• Code quality and security
• Documentation completeness
• Proper MCP protocol implementation
• No malicious code or vulnerabilities
• Clear licensing

After Approval

Once approved:

• Your server appears in marketplace listings
• Users can discover, install, and review it
• You can update the server by updating your repository
• You'll receive notifications about reviews and feedback

API Key Issues

Generation Issues

Installation Issues

Account Issues

Writing Effective Prompts

Code Quality

• Always test generated code thoroughly

  Run unit tests, integration tests, and manual testing before publishing

• Implement proper error handling

  Catch exceptions, validate inputs, and provide meaningful error messages

• Use environment variables for sensitive data

  Never hardcode API keys, passwords, or credentials

• Follow language-specific conventions

  Use linters, formatters, and style guides (PEP 8, ESLint, etc.)

Documentation

Security

Performance Optimization

• •Implement caching for frequently accessed data
• •Use connection pooling for database operations
• •Implement pagination for large datasets
• •Use async/await patterns to avoid blocking operations
• •Monitor and log performance metrics

Community Engagement