Taking over a system you did not build is normal work. It goes wrong when the handover is treated as a formality — a zip file, a password spreadsheet, and a promise that “it just runs”.
The inventory that matters
- Where does it actually run? Hosting account, server, container, someone’s personal cloud account. Find the invoice; the invoice never lies.
- Who owns the domain and the certificates? Registrar, expiry dates, and whether the renewal email goes to a person who still works there.
- Where is the source? Repository, branch that matches production, and how a deploy actually happens today.
- What are the backups? Not whether they exist — whether anyone has restored one, and how long it took.
- What talks to it? Payment providers, mail relays, analytics, and any integration whose credentials expire.
Do the inventory before you agree to a service level, not after. The first restore test is usually the moment you learn what you have really taken on.