Skip to content
Page 6 of 6

Reflection 3

How This Reflection Works

Your Facilitator will pull from these questions to lead a cohort-wide discussion. If your AI assistant is building during a Challenge and you have a free moment, these also make great team conversation starters while you wait.

What You Built

  • When an automated test catches a regression (a feature that broke when you added something new), what would have happened without that test? How long might the bug have gone unnoticed, and what's the cost of finding it later versus finding it immediately?
  • How does having a safety net of automated tests change the kinds of features you're willing to attempt? Would you take bigger swings at ambitious features if you knew the tests would catch you if something broke?
  • Is there a feature on your live URL right now that you're genuinely proud of? Pull it up - what makes it work?
  • As an application grows, it tends to get harder to change: new features take longer, and changes cause unexpected side effects. What causes that, and what's the signal that a codebase needs reorganizing before you add more to it? How do automated tests help (or not help) with this problem?

What You Practiced

  • Lesson 3 introduced the closed loop: write criteria, write a test, watch it fail, then let AI implement until the test passes. How does giving AI a failing test to work toward differ from giving it a description of what to build? Why might AI produce better results when it has a concrete, verifiable target?
  • Once you've established a workflow like the closed loop, how do you make it the default behavior instead of something you have to remember each time? What's the difference between a workflow you enforce manually and one that's encoded into your tools?
  • Lesson 3 described the "two-week cliff": the point where changes start silently breaking previously verified work. What causes that cliff, and how do automated tests change the equation? Without tests, how would you even know something broke?

How You Worked

  • When automated tests handle verification instead of manual review, how does a team's workflow change? Can you work faster, take bigger risks, or divide the work differently when a test suite has your back?
  • Think about accountability. When you're responsible for AI's output but can only spot-check it, how confident are you in what ships? How does having automated tests change that? What's the difference between "I looked at it and it seemed fine" and "the tests pass" as a basis for confidence?
  • Did anyone on your team try running the live URL on their phone or sharing it with someone outside the team? What was the reaction?

Looking Ahead

  • You're building one feature at a time: write criteria, generate tests, implement, verify, ship. It works, but you're still the bottleneck. What would change if you could delegate multiple stories in parallel and have AI work on them simultaneously?