It will never cease to amaze me that in the year of our Lord 2026, companies are still relying on screening out candidates with twee little "Leetcode"-style logic puzzles rather than anything that actually assesses engineering skills or knowledge. I have over 12 years in this industry, and I have shipped code that's running on the servers of Fortune 100 companies, major publications, and national professional sports teams.
And yet, at no point in my career has my success ever been measured by my ability to debug an overcomplicated FizzBuzz or dissect a RegEx while a 2-minute timer is ticking in my face; and furthermore, _with no access to resources or documentation_. Yes, if you tab away from the test window to check on the exact signature of some function because you don't have it committed to memory, you'll get a warning. The language they purport to use is well-known to have inconsistent method signatures and typing, so apparently you're expected to have the language manual down cold by rote; no actual assessment of thought process or skill.
So here's the big part: I can't do them as presented. My ADHD-riddled brain won't unclog long enough to walk the logic with a timer ticking in my face. And I learned a long time ago: that's not a reflection on me. As soon as the artificial constraints are removed, it's easy. It's a factor that's ONLY ever present in the artificial environs of these "assessments." And since there are a ton of people, maybe even newer developers, in these same shoes, I want to emphasize this: _you are better than this_. My own career is proof enough. Five or ten years ago, stuff like this would have smashed my confidence, but now it just makes me mad that companies are still out here doing this and offering no alternatives, keeping otherwise talented engineers outside of an arbitrary gate.
I'm going to keep looking, and you should too.