Ask HN: How do you get comfortable with shipping code you haven't reviewed?

7 points by fnimick 154 days ago | 12 comments

This is the advice I've gotten on how to adapt to AI driven development at breakneck speed - to the point of having AI tooling write and ship projects in languages the 'operator' doesn't even know. How do you get confidence in a workflow where e.g. a team of agents does development, another team of agents does code review and testing, and then it is shipped without a human ever verifying the implementation?

I hear stories of startup devs deploying 10-30k+ lines of code per day and that a single dev should now be able to build complete products that would ordinarily take engineer-years in under a month. Is this realistic? How do you learn to operate like this?

mattmanser 153 days ago |

You don't. Whoever's telling you those stories has a very long nose.

2rsf 153 days ago | |

100% agree with the "you don't", but I wouldn't be surprised if young startups or highly stressed teams delivering low-risk products will do just that and deliver unreviewed code

Sevii 154 days ago |

You get confidence in things by doing them. If you don't have experience doing something, you aren't going to be confident at it. Try vibe coding a few small projects. See how it works out. Try different ways of structuring your instructions to the 'agents'.

fnimick 154 days ago | |

Are there public examples of "good instruction" and an iteration process? I have tried and have not been very successful at getting Claude Code to generate correct code for medium sized projects or features.

Jeremy1026 153 days ago | | |

I had Claude write a piano webapp (https://webpiano.jcurcioconsulting.com) as a "let's see how this thing works" project. I was pleasantly surprised by the ease of it.

I actually just put together a write up showing my prompts and explaining what was generated after each, if you're interested at all https://jcurcioconsulting.com/posts/how-i-used-claude-code-t...

Sevii 153 days ago | | |

Anthropic has a short training course https://www.coursera.org/learn/claude-code-in-action. There isn't really a lot of best practices at this point because the technology has improved significantly in 2025.

LarryMade2 151 days ago |

I can see companies could get away with it from new coders who don't understand the repercussions of releasing )buggy code or know much of long-term maintenance.

You know, this sounds familiar - the big outsourcing boon to less expensive development in other countries, sure they can write the code, but it's either shipped buggy or takes a lot of management and hand-holding of the outsourced development teams to get everything right.

didgetmaster 153 days ago |

Sometimes it feels like there is an awful lot of software out there that shipped without much review. This was happening long before AI arrived on the scene.

Hard to tell if anyone was 'comfortable' with that.

binsquare 152 days ago |

The same way you learn to trust other dev's to do work.

I see ai as a tool, not a peer. I trust a peer when we aligned on the requirements of the project and where we want to go.

So the answer to how we get confidence in a workflow of agents to develop, to review and test without a human verifying the implementation?

I personally don't see me getting there.

muzani 153 days ago |

Likely a good deal of test coverage. At the far end of this is something like Facebook, which has everything monitored by A/B tests. If it breaks something that changes something serious, the alarm triggers. Move fast break things isn't a new way of doing things, so might as well pick up a framework that works.

chaidhat 153 days ago |

Hi, I’m running a 4-person startup based in Bangkok, Thailand and we differentiate code quality based on priority. We try to ship only clean code on master but when we talk with clients and they want a demo of a new product/feature, we use AI to rapidly create an MVP to see if it aligns with their needs. If they are happy, we then refine this MVP until we are happy with the code through manual review and refactoring, or we even rewrite it. We make sure our data is shaped correctly, hot paths are tested and things are well separated by domain (Domain driven design). DDD ensures us that if the code is shitty, only that part of the project is shitty. Only when the code is acceptable, we rebase to master. I try to let engineers talk to clients so that they learn the most from them first hand and then let them dictate smaller tasks for AI to do —- they are more product managery than what a typical engr would be ten years ago. Do you think this is a good approach? I’m also curious what other startups do too.

tldr we aren’t confident of the code we write quickly but we then take time to make sure we’re confident before we merge to master