07:31Sip and Sync: Super-charging GitHub Copilot with Model Context Protocol (MCP), with Krystal Folkes
From the creator
In this episode, Priyanka Vergadia invites Krystal Folkes back to showcase Model Context Protocol (MCP) - the new way to plug external tools into GitHub Copilot and make your AI agents truly extensible. Watch Krystal walk through enabling MCP in VS Code, wiring up a Playwright MCP server via npm, and running UI-automation commands right from the chat window. You’ll leave with a clear picture of how MCP can unify agents, back-end APIs, and developer workflows under one common “language.” Chapters 00:00 – Intro 00:28 – Why Model Context Protocol matters for AI agents 01:30 – MCP capabilities at a glance 02:10 – Installing VS Code Insiders and enabling MCP settings 03:05 – Copilot Chat agent mode explained 03:45 – Adding an MCP server: Playwright via npm 05:00 – Live demo – Navigating a site with Playwright commands 06:35 – Running custom Playwright scripts from the chat 07:40 – Imagining multi-agent travel and reservation scenarios 08:52 – Top three benefits of MCP for developers 09:42 – Closing thoughts and next steps What You’ll Learn -How MCP gives AI agents and back-end services a shared contract for communication -Setting up VS Code Insiders and turning on Copilot Chat “agent mode” -Installing an MCP server (Playwright example) and adding it to Copilot Chat -Driving browser automation, tests, and more through natural-language prompts -Real-world ideas for chaining multiple reservation or travel agents together -Productivity gains from keeping everything inside your IDE
Published 2025-06-10 · Watch on YouTube ↗
Availability and playback restrictions are controlled by the original channel.





