Run the MCP Client

What is Run the MCP Client?

Run the MCP Client is a Python client for testing a Model Context Protocol (MCP) filesystem server. It connects to a running server, lists available tools, and invokes them—defaulting to a write_file call that creates a test file. It is aimed at developers who want to verify their MCP server setup.

How to use Run the MCP Client?

Create a virtual environment, install dependencies (mcp, python-dotenv), and set up a .env file with the server path and allowed directory. Run python client.py to execute the default tool (write_file). Optionally pass a specific tool and arguments via CLI flags --tool and --args.

Key features of Run the MCP Client

• Connects to any MCP filesystem server.
• Lists all tools exposed by the server.
• Default tool call creates a test file.
• Configurable via .env or command‑line arguments.
• Supports testing multiple tools (list_directory, read_file, etc.).

Use cases of Run the MCP Client

• Verify that an MCP filesystem server is running correctly.
• Test different file operations (write, read, list) without writing custom code.
• Validate server‑client connectivity and tool availability.
• Learn how MCP clients interact with a server.

FAQ from Run the MCP Client

How do I install the dependencies?

Run pip install -r requirements.txt inside a Python virtual environment. The required packages are mcp and python-dotenv.

How do I change which tool the client calls?

Set the TOOL and ARGS environment variables in .env, or pass them as --tool and --args when running python client.py.

What does the client do by default?

If no tool is specified, it calls write_file to create test.txt with the content “Hello from Python the new standup MCP client!” in the allowed directory.

Do I need a server already running?

Yes. The client expects an MCP filesystem server (e.g., the Node.js server from the referenced repository) to be started and listening on the path specified in SERVER_PATH.