mcp man

mcp-serverman is a Python-based command-line tool to manage Claude MCP servers configuration with version control and profiling. It provides features to initialize a client configuration, list and manage multiple MCP servers, enable or disable specific servers or versions, and stage changes with version control. The companion mcp server capability allows Claude/LLMs to help configure and manage MCP servers for you. Core commands include initializing the client, listing servers, enabling/disabling/removing servers, saving changes with a message, and changing server versions. The tool also supports presets for quick configuration and multi-client workflows, enabling you to manage several configurations from a single interface.

To use it, first install the package, then run the mcp-serverman CLI directly in your terminal. Typical workflows include initializing the client, listing configured servers, enabling a server you wish to manage, and using the save and change commands to track versioned configurations. For advanced workflows, you can employ presets to quickly load predefined configurations or manage multiple clients and representatives of your MCP server configurations. The companion server feature helps you delegate configuration tasks to Claude/LLMs through the MCP server framework.

Prerequisites:

• Python 3.8+ (recommended)
• pip (comes with Python)
• Optional: virtual environment tool (venv) for isolation

Install from PyPI:

pip install mcp-serverman

Install from GitHub (latest debug/versioned branch):

pip install git+https://github.com/benhaotang/mcp-serverman.git

Verify installation:

mcp-serverman --version

Basic usage after installation can start right away by invoking the command:

mcp-serverman

If you plan to develop locally, you can install in editable mode from the repo root:

pip install -e .

Environment considerations:

• Ensure your Python environment has access to network and permits package imports.
• If using a virtual environment, activate it before installing.

Tips and common issues:

• Back up your MCP configuration before making changes, as recommended in the project notes.
• The companion MCP server feature requires appropriate client configuration and may depend on your specific environment and path settings.
• If you encounter permission issues on Windows, try running the terminal as an administrator or ensure Python and pip are installed with user-level access.
• For managing multiple servers or clients, explore the presets and multi-client commands to streamline workflows.
• Consult the manual for detailed usage beyond the CLI defaults (referenced in the README).