Demo agents look magical. Production agents look... fragile. Where does yours crack first?

Every demo I've sat through this year looked amazing. Clean inputs, happy path, a nice "it just works" moment. Then week one in production hits and it's a different animal.

Where I keep seeing the cracks:

- inputs nobody planned for (scanned PDFs, half-filled forms, emails in 3 languages)

- a tool call that times out and the agent just... keeps going

- edge cases that make up maybe 10% of volume but 80% of the headaches

Honestly the gap between demo and prod is the part of this job that keeps me humble. At AI Hive most of our time goes into that boring middle, not the model.

So for folks shipping agents right now: what broke first for you, and how long did it take to notice?