About mcp-client
mcp-client is a web application paired with a LangGraph agent. It is structured for developing an agent that connects to MCP servers and calls their tools, with a CopilotKit frontend.
Key features
- 01LangGraph-based agent
- 02Connect to MCP servers
- 03Call MCP tools
- 04CopilotKit frontend
- 05Run frontend and agent separately
Use cases
- 01Build an agent that invokes MCP tools
- 02Develop the frontend and agent independently
MCP support
The LangGraph agent connects to MCP servers and calls their tools. The project README does not describe server management or additional MCP capabilities.
Details
From the project READMEGetting Started
Set Up Environment Variables:
touch .env
Add the following inside .env at the root:
LANGSMITH_API_KEY=lsv2_...
Next, create another .env file inside the agent folder:
cd agent
touch .env
Add the following inside agent/.env:
OPENAI_API_KEY=sk-...
LANGSMITH_API_KEY=lsv2_...
Development
We recommend running the frontend and agent separately in different terminals to debug errors and logs:
# Terminal 1 - Frontend
pnpm run dev-frontend
# Terminal 2 - Agent
pnpm run dev-agent
Alternatively, you can run both services together with:
pnpm run dev
Then, open http://localhost:3000 in your browser.
Architecture
The codebase is split into two main parts:
/agentfolder – A LangGraph agent that connects to MCP servers and calls their tools./appfolder – A frontend application using CopilotKit for UI and state synchronization.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.