Investigating the belt bug

#corporate-primary-navigation { display: flex !important; } .site-header__menu-toggle { display: none !important; } .site-header--corporate .division-switch__dropdown-trigger { display: none !important; } @media (max-width: 900px) { .site-header--corporate { position: static !important; } .site-header--corporate .site-header__inner { display: flex !important; flex-wrap: wrap; padding-block: 0.75rem; } #corporate-primary-navigation { position: static !important; display: grid !important; grid-template-columns: 1fr; flex: 1 0 100%; width: 100% !important; max-height: none; overflow: visible; } .site-header--corporate .division-switch__dropdown-shell { display: block !important; } }

Back to devlogs

ACTIVED / DEVLOG

Belts look simple until each actor state stops agreeing with the next one. The current work is to log input, output and blockage so the exact jam point can be found.

The goal is not just to fix one isolated bug, but to make the system more observable for the next production layers.