MCP Chat with Claude
A typescript example for web app as the host connecting node mcp server
jiangyan
README
MCP Chat with Claude
A full-stack application demonstrating the integration of Model Context Protocol (MCP) with Anthropic's Claude LLM, providing an interactive chat interface that leverages MCP tools.
Project Overview
This project consists of two main components:
- MCP Server: A Node.js server implementing the Model Context Protocol that provides various tools for the LLM to use.
- Host App: A web application that serves as a chat interface and acts as the intermediary between the user, Claude AI, and the MCP server.
Features
- Interactive chat interface with Claude AI
- Two-panel UI showing conversation and tool execution details
- Dynamic tool discovery from MCP server
- Support for multiple tools with different parameters
- Proper handling of Claude's tool use via MCP
Project Structure
├── client/ # Host app & web client
│ ├── dist/ # Compiled TypeScript
│ ├── host-app.ts # Host application server
│ ├── chat.js # Frontend chat interface logic
│ ├── index.html # Main web interface
│ ├── styles.css # Styling for web interface
│ ├── package.json # Client dependencies
│ └── mcp-client-example.ts # Example MCP client for testing
│
├── server/ # MCP server
│ ├── dist/ # Compiled TypeScript
│ ├── mcp-server.ts # MCP server implementation
│ └── package.json # Server dependencies
│
├── .gitignore # Git ignore file
└── README.md # Project documentation
Prerequisites
- Node.js 17 or higher
- npm or yarn
- Anthropic API key
Installation
-
Clone the repository:
git clone https://github.com/yourusername/mcp-chat-with-claude.git cd mcp-chat-with-claude
-
Install dependencies for both client and server:
cd server npm install cd ../client npm install
-
Create
.env
file in the client directory with your Anthropic API key:ANTHROPIC_API_KEY=your-api-key-here MCP_SERVER_HOST=localhost MCP_SERVER_PORT=8000 MCP_SERVER_PATH=/sse
Setup and Running
Step 1: Start the MCP server
cd server
npm run build
npm start
The MCP server will start on port 8000 by default.
Step 2: Start the host app (web server)
cd client
npm run build
npm start
The web server will start on port 3000 by default.
Step 3: Access the chat interface
Open your browser and navigate to:
http://localhost:3000
Available MCP Tools
The MCP server provides the following tools that Claude can use:
-
get-todo: Get a todo item for a specific category
- Parameters:
category
: String (e.g., "life", "work", "family", "friends")
- Parameters:
-
get-plan: Get the overall plan
- Parameters: None
Additional Example
The project includes a standalone MCP client example (mcp-client-example.ts
) that demonstrates how to connect to the MCP server programmatically without the web interface.
To run this example:
cd client
npm run client
Development
MCP Server Development
To add a new tool to the MCP server:
- Open
server/mcp-server.ts
- Add a new tool definition following the existing pattern:
server.tool("tool-name", { param1: z.string() }, async ({ param1 }) => { // Tool implementation return { content: [{ type: "text", text: "Result" }] }; } );
- Rebuild and restart the server
Host App Development
The host app consists of:
- Backend (
host-app.ts
): Express server that communicates with both Claude and the MCP server - Frontend (
chat.js
,index.html
,styles.css
): Chat interface that communicates with the backend
Architecture
┌─────────────┐ ┌────────────────┐ ┌───────────────┐
│ │ │ │ │ │
│ Web UI │◄─────►│ Host App │◄─────►│ MCP Server │
│ (Browser) │ │ (Express) │ │ (Node.js) │
│ │ │ │ │ │
└─────────────┘ └───────┬────────┘ └───────────────┘
│
▼
┌───────────────┐
│ │
│ Claude API │
│ (Anthropic) │
│ │
└───────────────┘
Troubleshooting
-
Connection issues: Ensure both servers are running and check the console for error messages.
-
Tool not found: Make sure the MCP server is running and the tool names match exactly.
-
API key errors: Verify your Anthropic API key is correctly set in the
.env
file. -
Tool use errors: Check the response panel for detailed error information.
License
Acknowledgments
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
MCP Package Docs Server
Facilitates LLMs to efficiently access and fetch structured documentation for packages in Go, Python, and NPM, enhancing software development with multi-language support and performance optimization.
Claude Code MCP
An implementation of Claude Code as a Model Context Protocol server that enables using Claude's software engineering capabilities (code generation, editing, reviewing, and file operations) through the standardized MCP interface.
@kazuph/mcp-taskmanager
Model Context Protocol server for Task Management. This allows Claude Desktop (or any MCP client) to manage and execute tasks in a queue-based system.
Linear MCP Server
Enables interaction with Linear's API for managing issues, teams, and projects programmatically through the Model Context Protocol.
mermaid-mcp-server
A Model Context Protocol (MCP) server that converts Mermaid diagrams to PNG images.
Jira-Context-MCP
MCP server to provide Jira Tickets information to AI coding agents like Cursor

Linear MCP Server
A Model Context Protocol server that integrates with Linear's issue tracking system, allowing LLMs to create, update, search, and comment on Linear issues through natural language interactions.

Sequential Thinking MCP Server
This server facilitates structured problem-solving by breaking down complex issues into sequential steps, supporting revisions, and enabling multiple solution paths through full MCP integration.