Repository navigation
Home
From autocomplete to agent teams and production MCP servers.
Tip
📖 Read alongside the labs: the Agents & MCP Field Guide explains every concept in this course, with diagrams, analogies and a searchable glossary.
Work through the pages in the order listed below. Every page has ⬅️ Previous / Next ➡️ links at the bottom, and the sidebar shows the same order.
| Icon | Type | What to expect |
|---|---|---|
| 🧪 | Lab | Part A is guided step by step. Part B has less‑guided challenges with hints you open only when stuck |
| 📘 | Reading | Short handout with a quick self‑check |
| 🧗 | Challenge | Mostly unguided, with hints and a hidden reference solution |
| 🏆 | Capstone | Combine everything; graded with a rubric |
| 📎 | Reference | Look things up as needed |
Tip
Setup once: VS Code 1.140+, GitHub Copilot with AI credits, Node 24 LTS (≥ 22.19), Git, and a GitHub account. Use one Node version for everything.
| # | Step | Type | ⏱️ |
|---|---|---|---|
| 📘 | [[.11 · Timeline of Copilot in VS Code | .11 ‐ October 2026 ‐ GitHub Copilot in VS Code — Visual Timeline of Agents, Customization, and Automation]] | Reading |
| 📎 | [[.12 · Chat Symbols Cheat Sheet | .12 ‐ Github Copilot Chat “Symbols” Cheat Sheet]] | Reference |
| 📘 | [[.19 · Models & AI Credits | .19 ‐ Understanding Copilot Models & AI Credits]] | Reading |
| 📘 | [[.191 · Local Models (BYOK) | .191 ‐ Can You Run Your Own Local Models in GitHub Copilot (VS Code)?]] | Reading |
Module 1 companion readings (use alongside Labs 1 – 1.2):
| # | Step | Type | ⏱️ |
|---|---|---|---|
| 📘 | [[1.91 · Where Copilot Guidance Lives | 1.91 ‐ Reading ‐ Where Copilot Guidance Lives]] | Reading |
| 📘 | [[1.92 · When Repository Instructions Apply | 1.92 ‐ Reading ‐ When Repository Instructions Apply]] | Reading |
| 📘 | [[1.93 · Understanding the Baseline Output | 1.93 ‐ Reading ‐ Understanding the Baseline Output]] | Reading |
| 📘 | [[1.94 · The Plan Agent Explained | 1.94 ‐ Reading ‐ The Plan Agent Explained]] | Reading |
| 📎 | [[1.95 · Inside the Copilot System Prompt | 1.95 ‐ Reference ‐ Inside the Copilot System Prompt]] | Reference |
| # | Step | Type | ⏱️ |
|---|---|---|---|
| 📘 | [[2.01 · The Copilot Customization Map | 2.01 ‐ Reading ‐ The Copilot Customization Map]] | Reading |
| 🧪 | [[2.02 · Build an Agent Team with Custom Agents and Subagents | 2.02 ‐ Lab ‐ Build an Agent Team with Custom Agents and Subagents]] | Lab |
| 📘 | [[2.03 · Custom Agents vs Subagents | 2.03 ‐ Reading ‐ Custom Agents vs Subagents]] | Reading |
| 🧪 | [[2.04 · Watch Your Agents Work with Debug Logs | 2.04 ‐ Lab ‐ Watch Your Agents Work with Debug Logs]] | Lab |
| 📘 | [[2.05 · Why the Coder Agent Uses Context7 | 2.05 ‐ Reading ‐ Why the Coder Agent Uses Context7]] | Reading |
| 🧪 | [[2.06 · Shared Team Rules with an AGENTS.md File | 2.06 ‐ Lab ‐ Shared Team Rules with an AGENTS.md File]] | Lab |
| 📘 | [[2.07 · Agent Skills Explained | 2.07 ‐ Reading ‐ Agent Skills Explained]] | Reading |
| 🧪 | [[2.08 · Your First Skill with Scripts and Templates | 2.08 ‐ Lab ‐ Your First Skill with Scripts and Templates]] | Lab |
| 🧪 | [[2.09 · A Widget Skill for the Agent Team | 2.09 ‐ Lab ‐ A Widget Skill for the Agent Team]] | Lab |
| 🧪 | [[2.10 · Advanced Skill ‐ Telemetry Scenario Generator | 2.10 ‐ Lab ‐ Advanced Skill ‐ Telemetry Scenario Generator]] | Lab |
| 📘 | [[2.11 · Skills vs Tests and Mocks | 2.11 ‐ Reading ‐ Skills vs Tests and Mocks]] | Reading |
| 🧪 | [[2.12 · Self‐Healing Builds with Hooks | 2.12 ‐ Lab ‐ Self‐Healing Builds with Hooks]] | Lab |
| 📘 | [[2.13 · Local Agents vs Cloud Agents | 2.13 ‐ Reading ‐ Local Agents vs Cloud Agents]] | Reading |
| 🧪 | [[2.14 · Production Incident Fix with Copilot Cloud Agent | 2.14 ‐ Lab ‐ Production Incident Fix with Copilot Cloud Agent]] | Lab |
| 🧗 | [[2.15 · Ship a Feature with Your Own Agent Team | 2.15 ‐ Challenge ‐ Ship a Feature with Your Own Agent Team]] | Challenge |
| # | Step | Type | ⏱️ |
|---|---|---|---|
| 🧪 | [[3.00 · MCP Environment Setup (Prerequisite) | 3.00 ‐ Lab ‐ MCP Environment Setup (Prerequisite)]] | Lab |
| 📘 | [[3.01 · MCP Timeline and Why It Exists | 3.01 ‐ Reading ‐ MCP Timeline and Why It Exists]] | Reading |
| 📘 | [[3.02 · MCP Architecture | 3.02 ‐ Reading ‐ MCP Architecture]] | Reading |
| 📘 | [[3.03 · One Protocol, Many Transports | 3.03 ‐ Reading ‐ One Protocol, Many Transports]] | Reading |
| 📘 | [[3.04 · How VS Code Finds and Starts MCP Servers | 3.04 ‐ Reading ‐ How VS Code Finds and Starts MCP Servers]] | Reading |
| 🧪 | [[3.05 · MCP over HTTP with SQLite (Why Protocols Matter) | 3.05 ‐ Lab ‐ MCP over HTTP with SQLite (Why Protocols Matter)]] | Lab |
| 🧪 | [[3.06 · How Copilot Chooses an MCP Tool | 3.06 ‐ Lab ‐ How Copilot Chooses an MCP Tool]] | Lab |
| 📘 | [[3.07 · How Copilot Discovers the Database Schema | 3.07 ‐ Reading ‐ How Copilot Discovers the Database Schema]] | Reading |
| 🧪 | [[3.08 · Transport Swap from HTTP to stdio | 3.08 ‐ Lab ‐ Transport Swap from HTTP to stdio]] | Lab |
| 📘 | [[3.09 · What MCP Inspector Is For | 3.09 ‐ Reading ‐ What MCP Inspector Is For]] | Reading |
| 🧪 | [[3.10 · Protocol Forensics with MCP Inspector | 3.10 ‐ Lab ‐ Protocol Forensics with MCP Inspector]] | Lab |
| 🧪 | [[3.11 · The Dangerous Tool (MCP Security Boundaries) | 3.11 ‐ Lab ‐ The Dangerous Tool (MCP Security Boundaries)]] | Lab |
| 🧗 | [[3.12 · Replace Raw SQL with Safe Tools | 3.12 ‐ Challenge ‐ Replace Raw SQL with Safe Tools]] | Challenge |
| # | Step | Type | ⏱️ |
|---|---|---|---|
| 📘 | [[4.00 · MCP in Late 2026 ‐ What Changed | 4.00 ‐ Reading ‐ MCP in Late 2026 ‐ What Changed]] | Reading |
| 🧪 | [[4.01 · Beyond Tools with Resources, Prompts and Instructions | 4.01 ‐ Lab ‐ Beyond Tools with Resources, Prompts and Instructions]] | Lab |
| 🧪 | [[4.02 · Production Grade Tool Design | 4.02 ‐ Lab ‐ Production Grade Tool Design]] | Lab |
| 🧪 | [[4.03 · Human in the Loop with Elicitation | 4.03 ‐ Lab ‐ Human in the Loop with Elicitation]] | Lab |
| 🧪 | [[4.04 · Remote MCP Servers with Auth and Sandboxing | 4.04 ‐ Lab ‐ Remote MCP Servers with Auth and Sandboxing]] | Lab |
| 🧪 | [[4.05 · Testing MCP Servers in CI | 4.05 ‐ Lab ‐ Testing MCP Servers in CI]] | Lab |
| 🏆 | [[4.06 · Fleet Telemetry MCP Server for Your Agent Team | 4.06 ‐ Capstone ‐ Fleet Telemetry MCP Server for Your Agent Team]] | Capstone |
- Instructor ‐ Advanced Topics Plan: ten follow‑up modules with labs, challenges and a two‑day workshop schedule.
- Billing moved from premium requests to AI credits, and Edit mode is gone. All handouts are updated.
- Custom agents, skills and hooks match current VS Code (no extra settings needed, real hook schema, correct skill naming rules).
- MCP labs use the v2 SDK and were tested end to end with the MCP Inspector, in both the legacy (VS Code) and modern (2026‑07‑28) protocol eras.
- New Module 4 (Advanced MCP): resources and prompts, tool design, elicitation, remote servers with auth and sandboxing, testing in CI, and a capstone.
- Every lab now ends with less‑guided challenges and hints.