mcp_gradio_client

This MCP Gradio Client Proof of Concept provides a Gradio-based interface to interact with MCP servers over STDIO and SSE. The client automatically discovers tools exposed by configured MCP servers and presents them in the Gradio UI, allowing a language model to call tools through a standardized MCP protocol. You can run STDIO servers (launched automatically by the client as needed) and/or SSE servers (which must be started separately) and then load their tools into the interface for real-time usage. The UI supports dynamic tool loading, debugging mode, and a straightforward chat-style interaction with the model while invoking tools when appropriate.

Prerequisites:

• Python 3.12 or higher
• Node.js (for any optional Node-based tooling)
• uvicorn and/or uvx (depending on your STDIO server setup)
• OpenAI API Key for language model interaction (if using OpenAI models)

Steps:

1. Clone the repository (replace with the actual URL): git clone https://github.com/yourusername/mcp-gradio-client.git cd mcp-gradio-client

2. Create and activate a virtual environment:

   • Unix/macOS: python -m venv venv source venv/bin/activate
   • Windows: python -m venv .venv ..venv\Scripts\activate

3. Install dependencies: pip install -r requirements.txt

4. Set up environment variables (example): Create a .env file in the project root and add your API key references, e.g.: OPENAI_API_KEY=your_openai_api_key

5. Run the Gradio UI: python gradio_ui.py

Notes:

• STDIO servers defined in config.json are instantiated automatically by the Gradio client as needed.
• SSE servers must be started manually before launching the Gradio UI.
• If your setup uses different server command names (e.g., npx on Windows), adjust config.json accordingly.

Tips and notes:

• Ensure your config.json is present in the project root and follows the mcpServers structure shown in the README.
• For SSE servers, start the server prior to running the Gradio app and verify the URL (and any required headers) in your config.
• The Gradio UI provides a debug toggle; enable it to capture detailed logs for troubleshooting.
• If you encounter tool-loading delays, verify network connectivity to the SSE endpoint and that STDIO servers can be started with the configured commands.
• The client is designed to dynamically load tools from MCP servers; if a server exposes new tools, refresh or reload the interface to see them.