mcp-config-manager

MCP Configuration Manager provides a visual workflow for creating, editing, and syncing MCP server configurations without hand-editing JSON. The app offers a visual workspace to drag and drop MCP servers, a server discovery pane to browse the MCP marketplace, and a Monaco-powered editor for advanced JSON editing when needed. Use it to validate configurations in real time, auto-backup changes, and automatically sync server setups across your AI clients. The tool is designed to unify management across Claude Desktop, VS Code, Cursor, Kiro, Windsurf, Gemini, and other MCP clients, reducing configuration drift and speeding up deployments.

Within the visual workspace you can browse available MCP servers, install them with one click, and connect them to various AI clients. Each server card shows status, metrics, and configuration options. Real-time validation prevents broken setups, and automatic backups ensure you can rollback if a change introduces an issue. The built-in configuration templates let you apply predefined setups across projects or teams, while the hierarchical scope features support global, project, and user-level configurations for flexible governance.

Prerequisites:

• A macOS host with macOS-friendly tooling (the app is distributed as a macOS app)
• Administrative access to install apps from downloaded packages
• Optional: Docker if you prefer to run the server component in a container

Step-by-step install (macOS):

1. Download the latest macOS release (arm64 or intel) from the official releases page.
   • macOS Apple Silicon (M1/M2/M3): MCP.Configuration.Manager-0.1.8-arm64.dmg
   • macOS Intel: MCP.Configuration.Manager-0.1.8.dmg
2. Open the downloaded DMG and drag the MCP Configuration Manager app into Applications.
3. If you encounter a first-launch security warning, right-click the app in Applications and choose Open, then confirm to run.
4. Launch the application from Applications. On first run, it will scan for connected AI clients and set up basic workspace defaults.
5. Optionally, set up Docker-based server execution by installing Docker Desktop and ensuring the container image is accessible (if you plan to run via container).

If you prefer Docker without the macOS app:

• Ensure Docker Desktop is installed and running.
• Use the mcp_config via docker run as described in the mcp_config section to start the manager container.

Tips and common issues:

• Backups: The tool automatically creates backups before applying changes. If you need to rollback, use the rollback option in the Advanced/History panel.
• Environment variables: Configure MCP_CONFIG_PATH to control where configurations are stored; DEBUG can be enabled for verbose logs during troubleshooting.
• Validation: Real-time validation will flag JSON syntax errors or misconfigured server-to-client connections before they break your setup.
• Scope management: Leverage Global, Project, and User scopes to enforce consistent configurations across teams while allowing personal tweaks where needed.
• Connectivity: Ensure your AI clients (Claude Desktop, VS Code, Cursor, etc.) are discoverable by the manager, and that any required dependencies for the MCP servers are installed.
• Updates: When upgrading to a new vX.Y.Z release, review the changelog for breaking changes related to server templates or discovery surfaces.