Description
Master Model Context Protocol (MCP) from the ground up in this hands-on engineering tutorial. Instead of jumping straight into abstractions, we start by building a Python LLM application with memory and custom function calling, so you understand the exact bottleneck MCP solves. Before building STDIO and HTTP MCP servers, multi-server clients, and integrating the real-world GitHub MCP server.
Repo : https://github.com/telusko-aliens/telusko-youtube/tree/main/MCP-Youtube
Docs : https://docs.telusko.com/docs/mcp/getting-started
Check out our courses:
Agentic AI with Java : Live Cohort : https://go.telusko.com/agentic-ai-java-cohort-yt
Coupon: TELUSKO10 (10% Discount)
For More Queries WhatsApp or Call on : +919008963671
website : https://telusko.com/
Instagram: https://www.instagram.com/navinreddyofficial/
LinkedIn: https://in.linkedin.com/in/navinreddy20
Android App: https://go.telusko.com/get-app-playstore
iOS App: https://go.telusko.com/get-app-appstore
WhatsApp Community: https://go.telusko.com/whatsapp
Telusko Community: https://go.telusko.com/community
Discord: https://discord.gg/D8hWe9BqfF
00:00 - Recap
01:08 - Discussing AI Applications and LLM
04:07 - Exploring Models
12:13 - Building Your First LLM App in Python
20:06 - Adding Conversation Memory to the Chat App
25:04 - Why LLMs Need Tools (Explained)
28:58 - Implementing Custom Tools from Scratch
42:34 - Handling Multiple Tool Calls
46:03 - What is MCP (Model Context Protocol)?
58:33 - Building Your First MCP Server (STDIO)
1:02:23 - Connecting MCP Client (STDIO)
1:22:23 - Building MCP Server (HTTP)
1:24:05 - Connecting MCP Client (HTTP_STREAMABLE)
1:27:11 - Connecting an MCP Client to Multiple Servers
1:31:14 - Real-World Demo: Integrating the GitHub MCP Server
1:45:57 - Summary
