MCP -Creator

MCP Server Creator is an MCP server that helps you build and configure other MCP servers. It provides tools to dynamically generate server configurations, define tools with parameters and return types, manage resources (static and template-based), and generate runnable Python code for new MCP servers. You can interact with it either by running it directly as a server (via uvx) or by using its programmatic API to create and configure new servers, add tools, resources, and then generate or save the server code.

To use it as an MCP server, start it with uvx (the recommended method) and connect with an MCP client like Claude Desktop. You can also use the Python API to script server creation, tool addition, and code generation. The included example demonstrates creating a Weather Service, adding a tool to fetch weather data, introducing a resource template, and saving the generated server code to a file. This makes it straightforward to prototype, customize, and export complete MCP servers tailored to your needs.

Prerequisites:

• Python 3.8+ (as required by the project)
• pip (for Python packaging and dependencies)
• Optional: uvx (for running the server easily)

Installation steps:

1. Install the package from PyPI: pip install mcp-server-creator

2. Run the server using uvx (recommended): uvx mcp-server-creator

3. Alternatively, run as a module (if you prefer python -m): python -m mcp_server_creator

4. (Optional) Install development version: git clone https://github.com/GongRzhe/mcp-server-creator.git cd mcp-server-creator pip install -e .

5. Verify installation by listing available commands or by running a quick test script (as shown in the README example).

Tips and considerations:

• The server is designed to generate and export complete Python code for MCP servers, so you can iterate on designs quickly.
• Use the programmatic API to build complex servers with multiple tools and resources; the tool definitions support parameters, defaults, and asynchronous implementations.
• When using templates for resources, ensure your placeholders and parameter names align with the generated code to avoid runtime errors.
• If you encounter environment issues, verify that your Python environment is active and that uvx (if used) is installed and accessible in your PATH.
• The project requires Python 3.8+ and FastMCP at least version 0.1.0 as per the development notes.