How do you guys manage bug finding and the sorts
retained text
its very fustrating when you find out bugs on your vibe coded app, that cost you you quite a bit multiple times. I have implemented alot of things such as preview stage where me and agents can preview the features before they go live, test that run before every promote so all core vitals are checked before pushing to live servers. but still i get random costly bugs I could never think of unless i have the ability to remember all million lines of code. my most recent attempt is now making claude and codex keep spawning agents to think up of every single scenario of bugs that could happen untill it gets useless then run these large test every night at 3 am while my product is still actively being changed and then have similar agents spawned in everytime i make a new change and thinking about having an ai agent running every night to audit and check everything. and im also creating a visual ui of every single test with domain sorting so i can manually read them and get inspiration for other bugs that the ai hasn't thought off yet. Does anybody have good tips for this sort of bug hunting. also another thing ai loves to hard code things and make things not modular is there any method of auditing existing code bases to make sure it follows good practices like GRASP,GOF and the sorts i feel like when i ask claude to do that it doesn't provide anything meaningful. i cbb putting this in chatgpt to fix my typos.
document fetched 2026-10-04 00:34 · canonical https://reddit.com/r/ClaudeCode/comments/1wwp5by/how_do_you_guys_manage_bug_finding_and_the_sorts