Live, Snapshot, Cache: How a 3-Category Data Rule Prevented Three Real Bugs in 60 Days

A developer adopted a strict data classification rule in late April, requiring every stored value to be labeled as Live, Snapshot, or Cache — each with a distinct implementation constraint. The rule was born from an audit that uncovered weeks-long discrepancies in a totals column that had silently drifted without anyone questioning it. In three separate incidents over 60 days, the framework stopped a retroactive pricing script from overwriting historical enrollment fees, blocked a database migration that lacked a declared cache-refresh mechanism, and flagged a consumed redirect token before a flawed SMS campaign was sent to 53 contacts. Each near-miss was caught not by manual review alone, but by the discipline the classification system imposed at the point of design or commit. The author argues that naming the category of a stored value before creating it forces the right implementation questions early, closing the door on silent data drift before incidents occur.