The Balanced Engineer • Issue #37
Mastering GitHub's Copilot coding agent, measuring developer productivity using the SPACE model with AI, and training your attention muscles, with articles from Alexandra Lietzke, Lizzie Matusov, and Danny Kenny!
The Balanced Engineer Newsletter
Week of September 15, 2025 • Issue # 37
🔧 Technical Excellence
Building skills that last beyond any framework
GitHub Copilot coding agent 101: Getting started with agentic workflows on GitHub by Alexandra Lietzke on The GitHub Blog
Summary: This article is about GitHub's Copilot coding agent, which goes beyond traditional AI coding assistants by acting like an asynchronous teammate that can handle entire workflows from start to finish. Unlike regular AI tools that help you write code but leave you to manage branches, commits, and pull requests, coding agent can be assigned GitHub issues and will independently create branches, write code, run tests, and open pull requests for your review—all while operating in a secure, sandboxed environment with built-in protections. The agent is designed to tackle low-to-medium complexity tasks like bug fixes, refactoring, and improving test coverage, freeing you up to focus on the more interesting and strategic work. What makes this particularly compelling is that it integrates directly into GitHub's native workflow rather than requiring you to switch contexts, and with features like Model Context Protocol support, it can become even more context-aware and capable of handling complex development tasks autonomously.
Why this resonates: I'm preparing for my talk this week at CascadiaJS and want to do a shoutout to the Copilot coding agent, because it has been so critical in my thoughts behind my talk. As such, I've been reading more about it to see how other folks are using it. Using a coding agent in my day-to-day workflow has been groundbreaking for all of those "small tasks I already know how to solve that are too much effort to set up a dev environment for", as well as for targeted refactoring efforts. I know that AI Agents are supposedly going to be the future of development, so this seems like a muscle worth flexing. As such, I've been trying to do all that I can to smooth out my own workflow. My favorite part about the Copilot coding agent is that it's built into GitHub tools I already use on the regular: Issues, PRs, and Actions. It has made it easy to experiment with and to see how useful I can make it!
Tags: Artificial intelligence, Coding agents
🧠 Mental Models & Problem Solving
How we think about complex problems
How does the SPACE framework apply to AI's impact on engineering? from Lizzie Matusov from Research-Driven Engineering Leadership
Summary: This study shared by Lizzie Matusov studies how AI tools impact developer productivity across the SPACE framework (Satisfaction, Performance, Activity, Communication, Efficiency), or the widest, most accepted measurement of developer productivity. They found that the benefits of AI use go beyond speed, to higher job satisfaction and better ability to deliver business value. Research also revealed that AI adoption works best as a team habit, where developers on AI-supportive teams were 7x more likely to use AI daily and perceived stronger productivity gains both personally and team-wide. This shows that the impact of AI on developer experience is cultural, not just technical, requiring shared norms and open discussion to realize the full benefits across all dimensions of productivity.
Why this resonates: The biggest learnings from this were the differences in effect based on the team's shared norms. Adopting a pro-AI mental model, as a team, will make AI usage more useful overall. That's fascinating to me, and explains a lot of the opinions I've heard both for and against AI usage. I know based on chatting with a variety of engineers that the vast majority of folks that aren't using it seem to be in teams or orgs that don't support or encourage AI usage. Those are the folks that are most likely to miss out on new opportunities due to a lack of experience with AI in the future.
Tags: Developer productivity, Artificial intelligence
⚖️ Work-Life Integration
Sustainable practices for long-term success
Why your attention keeps slipping away (and how to get it back) by Danny Kenny from Big Think
Summary: This article explores how our attention works like a trainable muscle system rather than a fixed trait, drawing from NYU researcher Dr. Emily Balcetis's work on visual perception and goal achievement. The author argues that we're facing an unprecedented attention crisis where "continuous partial attention" leaves us perpetually scattered, making every task feel overwhelming because our focus simultaneously tracks multiple urgent priorities. The solution lies in training four distinct attention muscles: visual focus (narrowing your attention spotlight on specific targets), temporal connection (linking present actions to future outcomes), obstacle-processing (combining positive visualization with realistic planning), and cognitive flexibility (practicing unrelated focused activities). Rather than relying on willpower or time management alone, this research-backed approach treats attention as a skill that can be systematically developed, transforming how we experience and achieve our most important goals.
Why this resonates: I had a few insights while reading this article, which is why I felt it was worth sharing. I'm often thinking about how modern thinking is fragmented, in part because of the massive amount of competing demands on our attention and brains. Some things from this article stood out to me, like picking a specific visual target and how that can narrow your attention. I think this is why I feel so much more capable of focusing when I'm whiteboarding a problem. All that is in front of my is the whiteboard and the problem, so it's much easier to focus on finding a solution. Another mention was the obstacle-processing, which feels similar to how I work on problems: Poke a bunch of holes in the solution to find what I may be missing. By thinking about what may go wrong, or the obstacles, I tend to come up with better overall plans with those obstacles accommodated. Finally, when it comes to cognitive flexibility, I love doing some slightly different work to keep my brain sharp. Perhaps that's why I love the NYTimes Crossword so much!
Tags: Attention
What I've Been Building
A quick look at what I've been working on this week
- Overcommitted: Ep. 24 | Software Engineering Ethics and Social Media - How has social media changed the world, and what role have software engineers played in building it?
- Overcommitted Book Club: Check out the async Looks Good To Me book club!
Have comments or questions about this newsletter? Or just want to be internet friends? Send me an email at brittany@balancedengineer.com or reach out on Bluesky or LinkedIn.