36M devs just joined GitHub. Here are the 5 skills they use most. 💡
Did you know that every second a new developer joins GitHub?
By the time you’re finished reading this newsletter, hundreds more will have signed up.
As GitHub COO Kyle Daigle shared at this year’s GitHub Universe, more than 180 million developers now build on GitHub—with the equivalent of a city the size of San Francisco joining GitHub every single week.
And they're building big, driving record growth across the platform. In their first week, 80% use GitHub Copilot, showing how AI has become an integral part of what it means to be a developer.
Yes, GitHub has always been the home of software development. But it’s also helping millions of developers—from beginner to pro—build without limits and do the best work of their lives.
So, how can GitHub help you right now? Let’s dive into some of the biggest headlines from Universe and this year’s Octoverse Report, and uncover how to roll up your sleeves and get going. 👇
Skill #1: Orchestrate agents like a pro 🎼
At Universe, we introduced a new mission control for working with AI coding agents in GitHub. Instead of dealing with scattered projects across multiple tabs, you can now assign tasks, steer agent sessions, and track progress in one place.
Here’s how to try it:
- Get GitHub Copilot Pro. If you have a Pro or Pro+ plan, you already have access to Copilot coding agent. If you’re on GitHub Copilot Free, start your 30-day free trial of Pro to explore the coding agent and tasks.
- Open mission control. Visit github.com/copilot/agents. You’ll see a new mission control page to manage kick off and all your agent sessions in one place.
- Create a task. Click New Task. Then, describe what you need (for example, “Add pagination to the user list API”). You can even steer Copilot as it works by giving additional direction like, “include playwright tests.”
- Review and merge. The agent’s work appears as a pull request. Review, tweak, and merge just like you would any pull request.
✅ Pro tip: With the latest updates announced at Universe, Copilot code review now reviews pull requests automatically—flagging issues, suggesting fixes, and helping you ship cleaner code.
Skill #2: Make AI work for you 🛠️
The data says it all: 80% of new developers use Copilot in their first week, and over 1.1 million public repositories now embed LLM SDKs.
As we’ve previously talked about, AI is quickly becoming an essential part of software development. Just like you wouldn’t do math without a calculator, it’s starting to feel unthinkable to build without AI in 2025.
Here’s how to make AI a default part of your workflow:
- Get boilerplate tasks done faster in agent mode. Ask Copilot to write tests or have it help with a refactor. In agent mode, Copilot takes on multi-step tasks, considers all your working project files, and interacts with you to modify your codebase, not just suggest the next line of code.
- Lean on Copilot CLI for everyday commands. Prefer working from your terminal? Copilot CLI can do everything from multi-file editing, refactoring, and writing documentation, just like agent mode in IDEs. It also helps with shell commands and environment setup, so you can move fast right where you work.
- Review AI output, same as human code. Treat it as a draft, not gospel. Run tests, question decisions, and refine. That’s how you move from using AI to code faster to coding smarter. And with a Pro or Pro+ subscription, you can ask Copilot to review the code in your editor or assign Copilot to a pull request.
✅ Pro tip: Every GitHub user gets free, limited access to Copilot. Turn it on and start coding with AI today at github.com/copilot.
Skill #3: Build smarter collaboration loops 🔁
GitHub’s community just passed 180 million developers, driving record highs in commits, pull requests, and repositories this year. Collaboration isn’t just bigger, it’s faster, more global, and more interconnected than ever before.
To keep up, you need systems that scale as fast as the ecosystem itself.
Here’s how to sharpen your collaboration loop:
- Use pull request templates. Give contributors a checklist (e.g., tests, docs, security notes) so reviews stay consistent and efficient.
- Automate with GitHub Actions. Run tests, linters, or builds automatically when a pull request opens. Less waiting, more merging.
- Set up CODEOWNERS. Automatically assign reviewers based on file paths or components. No more manual tagging.
- Label for clarity. Use labels like ready-for-review or needs-fix to signal status across time zones and teams.
✅ Pro tip: Use pull request templates, CODEOWNERS, and GitHub Actions together to keep your review loop fast and consistent. That way you fix the process, not just the bug.
Skill #4: Make your stack AI-ready with Typescript 🤖
For the first time ever, TypeScript is the most-used language on GitHub, surpassing Python and JavaScript.
Why? Because typed systems give AI tools clearer instructions, safer outputs, and fewer hallucinations. In short, types make AI collaboration easier.
If you want to future-proof your codebase for AI-assisted development, start here:
- Add types to existing projects. In JavaScript, rename .js files to .ts and start small. Add types to utility functions first. In Python, use mypy or pyright to gradually enforce type hints.
- Let Copilot help with conversion. Ask Copilot: “Convert this file to TypeScript with proper interfaces” or “Add Python type hints and fix any type errors.” You’ll get solid scaffolding fast, and you can fine-tune the rest. Learn more on using Copilot to translate code to a different programming language.
- Adopt typed libraries. Choose frameworks that fully support static typing (e.g., Next.js, FastAPI, etc.). The more type data your AI tools see, the better their suggestions become.
- Enforce typing in CI. Add a GitHub Action to fail builds when types don’t match. It keeps your AI agents and teammates aligned.
✅ Pro tip: Strongly typed code prevents bugs and trains your AI collaborators to understand your architecture.
Skill #5: Collaborate across borders 🌎
This year, India added more than 5 million developers to GitHub, the largest single-year increase of any country. Emerging communities across APAC, LATAM, and Africa are also contributing code at record speed, reshaping how and where software gets built.
That means your next collaborator might live halfway across the world, and your workflow should make it easy for them to jump in.
Here’s how to make your projects globally friendly:
- Write a world-class README. Explain what your project does, how to install it, and how to contribute. Keep instructions clear, short, and free of local jargon.
- Add a CONTRIBUTING.md file. Spell out contribution rules (e.g., branch naming, code style, pull request etiquette). This reduces friction for first-time contributors.
- Make onboarding instant with GitHub Codespaces. Add a .devcontainer.json file so anyone can spin up your full dev environment in seconds.
- Use inclusive communication. Labels like good first issue and time-zone-friendly async reviews make open source work feel accessible to everyone.
✅ Pro tip: Turn on GitHub’s community standards checklist (in your repository’s Insights tab) to see what you’re missing for global collaboration readiness.
The bottom line
From orchestrating your first agent to collaborating across continents, GitHub helps you not only keep up but stand out in the world’s fastest-moving developer ecosystem.
It’s no wonder that a new developer joins GitHub every second. If you’re one of the new ones, welcome. If you’ve been around a while, we’re so happy you’re here. No matter who you are or where you’re from, GitHub is the best place to build what’s next, and honestly, we’re just getting started. Every commit, pull request, agent task, issue, and action adds up to something bigger: a future we’re all building together. 💫
More GitHub goodness:
🕶️ Explore the GitHub Shop. New merch just dropped, including the GitHub Copilot Amazeball.
🔥 Subscribe to our developer newsletter. Discover tips and tricks to supercharge your development.
🐙 Join our team. From engineers to writers, we’re always looking for the next great talent.
❤️ Sharing is caring. Repost this newsletter to your network.
✨ This newsletter was written and produced by Gwen Davis. ✨
I help businesses build innovative web solutions that are fast, scalable, and user-focused.
1wAI has truly changed the way we write code
Co-Founder & Tech lover🔗AI and Machine Learning| Innovation with Scalable, High-Performance Software Automations and Model Training Solutions. ⚙️RAG ✔️LLM ✔️ML ✔️CV ✔️DL
1w✔️
Educator, Data Analyst, and Creative Learning Specialist
1wWonderful insights, Mr. Gwen Devis. The news about GitHub’s rapid growth is truly exciting, and your analysis on why developers favor TypeScript over Python or JavaScript is exceptionally well explained.
Computer science | Software Development | AI & ML | Research | Academic Writing
1wWow! Git hub is really something
Student at DEDAN KIMATHI UNIVERSITY OF TECHNOLOGY (DeKUT)
1wHer am new so can you tell what do you do in here