Do you trust AI to push code to production?
by•
AI still makes mistakes when coding.
However, for simple fixes or features do you bother switching branches and testing locally before creating a PR or pushing to production? Or do you just ask Claude for a fix, review quickly, then push?
I saw an interview with Peter Steinberger (creator of Openclaw). Where he mentions he always pushes to main and almost entirely vibe codes. If you look at his contributions, you see how fast he ships.
Do devs need to be more trusting?
15 views
Replies
vibecoder.date
Trust but verify.
AI code velocity gives us the ability to actually have time to test. Why so few choose to do so is beyond me.
Your code needs 3 to do things:
A) Fail predictably and loudly
B) Log events with sufficient detail
C) Be reliable in adverse conditions
This means that your best investment is a testing environment and the ability to simulate these conditions.
Yous tart from the general to the specific.
End to end flows.
Integration
then unit.
Cover end to end first, guarantee functional requirements
@build_with_aj Completely agree, great take. How do you set up your testing environment? Do you lean more on third-party monitoring or internal logging?
vibecoder.date
@masebuilds I prefer to run my own internal stuff.
1st: local dev env, make sure it runs on local host.
2nd: staging replication as close as possible,
3rd: prod
It depends a lot of what tools you use.
I also run linters and other static analysis tools, they save time.
For me, AI is a strong assistant, not a decision-maker. I use it for quick fixes but I still test locally before pushing. I move fast but I don't skip safety. Shipping quickly feels good but breaking production feels worse. I trust AI, just not blindly.