Show HN: Claude Code guards built from my agent's repeated mistakes (github.com)

I use Claude Code everyday in my project, claude code wrote codes and I reviewed them.

Sometimes it repeats same mistakes even if I wrote it on Claude.md

I want to make something which can check and block it by itself

so I build seven small checkers they also have each testers which are in repo

and English is not my mother tongue

What mistakes did your agent keep repeating? I want to hear yours