// Visual Studio Code

A model-packed week — Chris walks through new releases including GPT-4.1, Gemini 2.5 Pro, O3, O4 mini, and Meta Llama 4 in Copilot and GitHub Models. He demos the strategy trend radar app, tries the new Codespaces agent mode button, and builds a GitHub Actions workflow that summarizes the changelog into a newsletter using GitHub Models.

Chris explains what Model Context Protocol (MCP) is and demos the new GitHub MCP server alongside Copilot Agent Mode in VS Code. The stream covers creating repos and issues through MCP tools, then builds a strategy visualization app from a screenshot using agent mode, vision, and the Playwright and sequential thinking MCP servers.

Extending Rock Paper Scissors to include Lizard and Spock using GitHub Copilot Vision to interpret a game diagram. Also covers Claude 3.7 Sonnet in Copilot, Dependabot version updates for Go modules, GitHub Codespaces for testing, and repository branch protection settings.

Building a Rock Paper Scissors game in Go for the gh-game GitHub CLI extension using GitHub Copilot Agent Mode. Explores the new GPT-4o code completion model, Next Edit Suggestions, and drives test coverage from 42% to 100% using Copilot inline chat and agent workflows.

Building a Tic Tac Toe game in Go for the gh-game GitHub CLI extension using GitHub Copilot Agent Mode. Also covers setting up CodeQL code scanning, fixing GitHub Actions workflow permissions, and a demo of GitHub Spark for creating micro apps from natural language prompts.

Chris introduces the GitHub Skyline CLI extension, a Go-based tool that generates 3D STL models and ASCII art visualizations of your GitHub contribution history. The stream covers the gh skyline command, 3D printing contribution graphs, and live refactoring of the Go codebase using GitHub Copilot Edits to split a monolithic main.go into modular components.

Builds Conway's Game of Life end-to-end using GitHub Copilot Chat and Copilot Edits in Visual Studio Code, demonstrating a realistic iterative AI-assisted development workflow. The video covers model selection (including Claude Sonnet 3.5), prompt engineering best practices, refactoring from a single-file prototype to separate HTML, CSS, and JavaScript files using Copilot Edits, and enforcing team coding standards with a Copilot instructions file. It also shows inline chat slash commands (/doc, /explain), Copilot Extensions, README generation, and deploying to GitHub Pages via a GitHub Actions workflow.

Demonstrates the o1-preview reasoning model integrated into GitHub Copilot Chat in Visual Studio Code, available to approved users via the chat model picker alongside GPT-4o and o1 mini. The video builds a JavaScript maze generator that supports keyboard navigation, BFS and DFS solving algorithms, and visual path rendering—using o1-preview's extended internal reasoning to satisfy complex multi-requirement prompts in fewer iterations than standard models.

Chris and Matt pick up inside GitHub Codespaces, getting the Player State service running locally on the first attempt thanks to Dapr's environment-variable secret store and Codespaces secrets. They then wire up the World Events Engine with Azure Storage Queue bindings and Table Storage state before hitting a stubborn .NET minimal API routing bug — becoming the cliffhanger that sets up the next session.

Chris and Matt solve a recurring ToolUp Days pain point — inconsistent developer environments — by setting up GitHub Codespaces with a custom dev container. They extend a community .NET template to include Go, configure Dapr component files for local Azure Storage access, and leverage Codespaces secrets as injected environment variables to avoid storing credentials in source code. By the end of the session they have a fully reproducible, cloud-hosted development environment that spins up in seconds.