Create a .env file with the following content (refer to the .env.example file for guidance):

plaintext APIFY_TOKEN=YOUR_APIFY_TOKEN LLM_PROVIDER_API_KEY=YOUR_API_KEY

Default values for settings such as mcpSseUrl, systemPrompt, and others are defined in the const.ts file. You can adjust these as needed for your development.

Run the client locally

bash npm start

Navigate to http://localhost:3000 in your browser to interact with the MCP server.

Happy chatting with Apify Actors!

ⓘ Limitations and feedback

The client does not support all MCP features, such as Prompts and Resource. Also, it does not store the conversation, so refreshing the page will clear the chat history.

References

Model Context Protocol
Apify Actors MCP Server
Pay-per-event pricing model
What are AI Agents?
What is MCP and why does it matter?

1e:T1968,

🧠 MCP Client TypeScript

A robust TypeScript-based tool-calling agent framework that integrates OpenAI's GPT-4o with MCP Tool Servers, featuring real-time Chain-of-Thought reasoning via Server-Sent Events (SSE).

🚀 Features

🔗 Tool-Calling via MCP Protocol - Seamless integration with any MCP-compatible tool server
🧠 ReAct Planning Loop - Structured reasoning using Plan/Thought/Action/Observation/Final Answer
💬 Streaming with SSE - Real-time visibility into agent thought processes
⚙️ Dynamic System Prompting - Automatic prompt configuration based on available tools
🔧 Easily Extendable - Simple configuration via mcp-config.json
🔄 Session Management - Maintain context across multiple interactions

📦 Project Structure

.
├── index.ts                 # Main entrypoint
├── config/
│   └── mcp-config.json      # MCP tool server configurations
├── public/
│   └── index.html           # Basic frontend for testing SSE
├── src/
│   ├── server.ts            # Express application logic
│   ├── routes/
│   │   └── sseRoutes.ts     # Routes for /stream-sse and /clear-session
│   └── services/
│       └── mcpClient.ts     # Core logic (sessions, OpenAI integration, MCP tools)

⚙️ Setup Instructions

Prerequisites

Node.js (v16+)
npm or yarn
OpenAI API key

1. Install Dependencies

npm install

2. Environment Configuration

Create a .env file in the root directory:

OPENAI_API_KEY=sk-...
PORT=3000                # Optional, defaults to 3000

3. Configure Tool Servers

Edit config/mcp-config.json to define your MCP tool servers:

{
  "mcpServers": {
    "bsc-mcp": {
      "args": ["./mcp/bsc-tool-server.js"]
    },
    "weathr-mcp": {
      "args": ["./mcp/weather-tool-server.js"]
    }
  }
}

Each server should be an MCP-compatible program that communicates via stdin/stdout.

🛠️ Build & Run

Production Build

npm run build

This command:

Compiles TypeScript to JavaScript
Copies config/ and

mysql-kusto-mcp

tester-mcp-client

mcp-creation

mcp-client

mcp-client

academy-mcp-client

mcp-client-with-openai

Kopilot-MCP

mcp-client-mqtt

SuperMapMCP

MCP-Server-and-client-Implementation-on-Linux