ultimate_mcp_client

Ultimate MCP Client is a comprehensive, asynchronous interface for interacting with MCP servers and tooling. It provides both a reactive Web UI and a feature-rich CLI to manage server connections, run tools, and explore conversation graphs with persistent state. You can connect to servers over stdio, SSE, or streaming-http transports, discover local and remote MCP servers, and leverage built-in observability features like OpenTelemetry metrics and traces. The client supports real-time streaming of AI responses, dynamic tool routing, and direct tool execution, enabling complex, stateful interactions with external capabilities and data sources.

To get started, install the required Python package (via your preferred method), configure the MCP client to point at your target MCP server, and run the client with the command above. Once running, you can use the CLI commands to connect to servers, start conversations, and issue tool calls. The Web UI offers a visual interface with live chat streaming, tool management modals, and a branching conversation view, while the CLI provides a powerful Textual UI dashboard and rich Markdown rendering for scripting and automation.

Prerequisites:

• Python 3.13+ installed on your system
• Access to the internet to install dependencies
• Optional: a virtual environment manager (venv/conda) if you prefer isolated environments

Step-by-step installation:

1. Create and activate a virtual environment (optional but recommended): python -m venv venv source venv/bin/activate # on Unix/macOS .\venv\Scripts\activate # on Windows

2. Upgrade pip: python -m pip install --upgrade pip

3. Install the Ultimate MCP Client package from PyPI or directly from GitHub:

   • From PyPI (if published): pip install ultimate-mcp-client
   • From GitHub (latest source): pip install git+https://github.com/Dicklesworthstone/ultimate_mcp_client.git

4. Verify installation by running the module (this will use the default config and run the server client): python -m ultimate_mcp_client --help

5. Start the MCP client with the recommended entrypoint (configured in mcp_config): python -m ultimate_mcp_client

Note:

• If you maintain a local configuration, you can adjust the mcp_config in your environment or place a config file that the client reads on startup.

Tips and notes:

• Ensure Python 3.13+ is installed and available in your PATH.
• The client supports multiple transports (stdio, sse, streaming-http); ensure your target MCP server is reachable via the intended transport.
• For production deployments, consider running the client behind a process manager (e.g., systemd, PM2) to ensure resilience and auto-restart.
• Environment variables can be used to customize behavior (e.g., endpoints, timeouts, and logging levels). Placeholder env map is available in mcp_config for your customization.
• If you encounter stdout corruption with multiple stdio servers, use the Safe IO configurations described in the project docs to isolate output streams.
• The client exposes both CLI and Web UI modes; for automation, the CLI (Typer-based) is ideal, while the Web UI is best for interactive exploration and monitoring.