The horror of reading legacy code written by a developer who left in 2014

We have all been there. You open a repository that hasn't been touched in a decade. No documentation. No architecture notes. Just variables named "temp1", "data2", and a 4,000-line functions file.
You change a single string, and suddenly three completely unrelated microservices collapse in production.
Reading legacy code feels less like engineering and more like doing historical archaeology while blindfolded in a minefield. It takes a specific type of mental resilience to survive it daily without losing your sanity.
How do you survive your current legacy codebase? Do you actively try to refactor it step-by-step, or do you just pray the server doesn't restart during your shift? Let's discuss in the comments.
P.S. If you survived today's debugging session, check out my profile bio for the ultimate badge of honor.