Aug. 12, 2026
Judging by LinkedIn and X, no one is writing code by typing anymore. Everyone's using AI coding tools like Claude Code, Codex, or Cursor. Lately, things have gone even further. Now there's a heated debate about whether one should still read the code produced by the AI tools. Does that mean that clean code is dead?
July 8, 2026
The AI coding output can vary from "That's a pure AI slop!" to "That looks great, let's ship it!". Let's take a look at 5 tips to help you consistently produce better output.
July 6, 2026
Automated testing is boring to software engineers. Everyone would like to build something that's more complex, feature-rich, or faster. So why do I care about it so much?
June 17, 2026
AI agents are really great, but they don't always follow all the rules. This article shows how to improve the feedback loop for AI agents using a custom linter, thereby improving the quality of the output.
June 9, 2026
How the auto dependency update broke production, and what we did to prevent that from happening in the future
June 2, 2026
We're moving faster than ever. Code is changing, APIs are changing. It's all great until international breaking API changes are deployed to production. Learn how to prevent this from happening.
May 18, 2026
Supply-chain attacks are becoming increasingly frequent. Dependency cooldown is a very simple yet effective approach to protect your codebase against them. Learn how to apply that inside your Python projects.
April 14, 2026
CheatsheetAI code review sounds great in theory. In practice, most teams just end up with more noise in their merge requests. In this article, I share 3 steps that made AI code review actually useful in my team and also reasonably priced (under $1.50 per review).
April 8, 2026
AI made coding faster. But faster code without a fast feedback loop is just faster chaos. If we can code faster, we need everything around it to keep up: pipelines, testing, deployments, and monitoring. In this article, we break down a Python development workflow designed for shipping reliably in the AI era. Four pipelines. Three environments. Small MRs, you're willing to throw away. Deployments that are non-events. This is the first in a series where we'll dig into each step - from automated pipelines to feature flags to monitoring that catches issues before your customers do.
March 31, 2026
CheatsheetMost Python test suites have the same problem: they cost more trouble than they save. CI jobs that take forever, tests that break on every refactor, flaky pipelines you keep restarting. After years of writing (and deleting) tests, I've landed on 7 qualities that separate tests worth keeping from tests that just slow you down.