MCP Space provides a no-code approach to building and deploying AI-powered MCP (Model Context Protocol) servers. The platform combines a Next.js frontend for visual design and an AI-driven Google ADK backend that generates TypeScript MCP server implementations. Use the frontend to define tools, agents, and workflows, then rely on the ADK backend to converse with you, gather requirements, and automatically generate compliant MCP server code. Once generated, you can configure environment variables and test the server before deploying it to Cloudflare Workers with a single-click deployment workflow.

Prerequisites:

• Node.js 18+ (for the frontend) or simply follow the ADK-driven workflow if you’re using the generated backend
• Python 3.9+ (for the Google ADK backend)
• pip (Python package installer)
• A Google Cloud account with Gemini API access
• Supabase account (for storage and auth in the full platform)

Installation steps:

1. Clone the repository git clone https://github.com/tharuneshwar-s/mcp-space.git cd mcp-space

2. Install and run the Frontend (optional for local UI access) cd frontend npm install cp .env.local.example .env.local

   Edit .env.local with required API keys

   npm run dev

   Frontend is available at http://localhost:3000

3. Install and run the Google ADK Backend cd ../google-adk pip install -r requirements.txt

   Create a .env file with necessary API keys and DB configuration

   adk web --session_db_url="postgresql+psycopg2://{DB_USERNAME}:{DB_PASSWORD}@{DB_HOST}:{DB_PORT}/{DB_NAME}" --allow_origins="*"

   The ADK server will be available at http://localhost:8080

4. Optional: configure the MCP server for deployment

   • Ensure environment variables (API keys, DB credentials) are set
   • Review the generated MCP server code under the appropriate project sections

5. Deploy

   • Follow the platform’s one-click Cloudflare Workers deployment flow from the UI once you have an MCP server generated

Note: The exact commands may vary if you generate a different MCP server shape. The repository includes both a frontend UI and a Google ADK backend to support the end-to-end workflow.

Tips and common issues:

• Ensure your environment variables are properly set in .env.local (frontend) and .env (ADK) before starting services.
• The ADK backend uses Gemini API; ensure your Gemini credentials and access are configured in the ADK environment.
• For deployment, you can test locally with the provided URLs before pushing to Cloudflare Workers.
• If you encounter CORS issues during local testing, verify the allow_origins setting in the ADK start command.
• Supabase stores project data; keep your database credentials secure and rotate them periodically.
• When generating MCP servers via the ADK, review the generated TypeScript server to ensure environment variables and keys are wired correctly for deployment.