Langchain MCP Client

Clone the LangChain MCP Client repository from GitHub to your local machine.

git clone https://github.com/guinacio/langchain-mcp-client.git
cd langchain-mcp-client

Use uv to create a virtual environment and install all required packages including optional extras. Alternatively, use pip with a standard venv.

# Using uv (recommended)
uv sync --all-extras --dev

# Or using pip
python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate
pip install -r requirements.txt

Start the application. It will open automatically in your browser at http://localhost:8501.

# With uv
uv run streamlit run app.py

# With pip/venv (venv activated)
streamlit run app.py

In the Streamlit sidebar, select your LLM provider (OpenAI, Anthropic, Google, or Ollama) and paste in your API key. Set temperature, max tokens, and timeout to match your use case. No environment variable files are needed — keys are entered directly in the UI.

In the sidebar, add one or more MCP server URLs (for SSE transport, e.g., http://localhost:8000/sse). You can test connectivity using the included sample weather server.

# Run the bundled sample weather server
uv run python weather_server.py
# Then enter http://localhost:8000/sse in the app

Once connected, available MCP tools appear in the sidebar. You can invoke them through the chat interface using natural language, or use the individual tool-testing panel to call tools directly with custom JSON parameters and inspect raw responses.