Most Codebases Aren't Really Ready For Agents
Agents can write code faster, but that doesn't mean you can ship changes safely at scale. Craig Sturgis and Justin Lacey discuss the readiness flags to look for before you set agents loose on a product people pay for.
Most codebases I've worked on aren't even close to ready for agents to ship changes safely at scale. Luckily, our best practices still hold up: regression testing, clean architecture, and attention to detail on changes.
Vibe coding an idea with no users is a very different thing from running agents against a product people pay for. We get into the readiness flags I look for, why building a regression suite went from a two-year project to something feasible in weeks, and the three things that matter most on every single change. My conversation with Justin Lacey (~8 minutes).
Related insights
Cheap Agent Attacks Are Coming
The cost of attacking software systems is dropping fast. Craig Sturgis and Justin Lacey discuss the practical near-term concerns after the Hugging Face incident.
You're Going to Need to Ship a Fix in Minutes, Not Days
Agents are making it cheap for attackers to find unlocked doors. The window to invest in CI, test coverage, and fast safe rollback is now.
GitHub Is on the Struggle Bus Again
GitHub is absorbing a tsunami of AI generated code, and the outages keep coming. I've never been more open to an agent native alternative.
Get More Like This
Follow along as I build and share what I learn
Found this helpful? Share it with your network!