Loading video player...
In this deep dive tutorial, you'll learn how to turn a basic Microsoft Copilot agent into a powerful agentic solution using VS Code, GitHub Copilot, Copilot Studio, and Work IQ. We start with a simple agent built in the Microsoft 365 Copilot app and progressively evolve it into an advanced autonomous Copilot agent capable of handling real workplace automation. By the end of this video, you’ll learn how to create an agent that can autonomously: • Detect and answer emails from people you select • Save email attachments to your OneDrive • Edit documents based on instructions in the email and work context • Generate a professional email reply and attach the revised document • Notify you in Microsoft Teams of actions taken Here are the video Chapters: 0:00 Starting with Microsoft’s Pro Code Tools 0:07 Reusing What We Already Built: Boss Brain in Agent Builder 0:39 Copying to Copilot Studio 1:36 Changing LLM in Copilot Studio 2:59 Adding MCP Servers in Copilot Studio 5:55 Adding a Trigger to the Agent Autonomous in Copilot Studio 6:56 Starting in VS Code 7:14 Setup of VS Code & GitHub Copilot 8:39 Cloning Copilot Studio Agent in VS Code 9:53 Agent Structure in VS Code 10:42 GitHub Copilot Chat Overview 11:47 Copilot Studio Skills add-on 12:12 Initializing your agent /init 14:42 Synching between Copilot Studio and Local VS Code Agent 15:34 Planning Agent changes /plan 15:47 What we want the Revised Agent To Do 17:10 Selecting LLM for GitHub Copilot Chat 20:09 Executing Plan of Agent Changes 21:54 Checking the changes in Copilot Studio 22:53 Tips + Testing agent in Copilot Studio 25:39 Demo of Autonomous Agent modified in VS Code & GitHub Copilot Chat 27:12 Lessons Learned What You'll Learn • How to move agents from Copilot Agent Builder to Copilot Studio • How Work IQ gives agents context across files, emails, and meetings • How to add Model Context Protocol (MCP) tools for Teams, Outlook, Word, and OneDrive • How to clone and edit agents in VS Code • How to use GitHub Copilot Chat in Agent Mode to design complex workflows • How Copilot can generate entire automation pipelines from a simple plan • How to test and deploy autonomous Copilot agents Tools Covered • Microsoft Copilot Studio • Microsoft 365 Copilot • GitHub Copilot Chat • Visual Studio Code • Work IQ • Model Context Protocol (MCP) About the Channel My name is Shervin Shaffie, a Principal Copilot Technical Specialist at Microsoft, where I work with global organizations deploying Microsoft AI technologies including Microsoft 365 Copilot, Copilot Studio, and enterprise AI agents. On this channel, I break down real-world AI workflows, Copilot agents, and automation scenarios to help developers, architects, and technical professionals understand how AI can transform everyday work. If you're interested in building practical AI agents for the workplace, you're in the right place. #CopilotStudio #MicrosoftCopilot #AIAgents #GitHubCopilot #VSCode #MicrosoftAI #AutonomousAgents *SUPPORT THIS CHANNEL ✅* https://www.youtube.com/channel/UCAeJheOWuUrs00ZPnUaOA0A/join FOLLOW ME 📸 ► Podcasts: https://rss.com/podcasts/sherv/ ► LinkedIn: https://www.linkedin.com/in/sherv/ ► Website: https://www.cosi.pro OTHER VIDEO SERIES 🎞️ ► 🤖: https://www.youtube.com/playlist?list=PL7gs6v-vK75qokJdj3fz6pMUWRZiTkjzr ► 👾: https://www.youtube.com/playlist?list=PL7gs6v-vK75pzr9Zy2WJUCea7NVs58d1E ► 💾: https://youtube.com/playlist?list=PL7gs6v-vK75p_KsgSJ4RSeaKE5-dwfI9x&si=0Ibnbr-myUiFhbkd This video represents the opinions of Shervin Shaffie and his guests to the video and does not reflect the opinions of Microsoft, affiliates, or the companies with which this video's participants are affiliated with. *Thank You All!!!* © 2020-2026 Collaboration Simplified. All rights reserved.