How would you prepare realistic test data without copying customer secrets?

Instruction: Give a concrete data and credential plan for the test suite and its CI artifacts.

Context:

A fictional account-management suite needs linked users, invoices, expired sessions, Unicode names, long addresses and missing optional fields. A developer proposes copying a recent customer database into CI because random fixtures have missed edge cases. CI traces and screenshots are retained for 14 days and can contain form data or request headers. The team can create synthetic records and issue limited test-environment credentials.

Updated

Official answer available

Read the opening below, then unlock the full answer and practical guidance.

I would build a small synthetic dataset that explicitly covers those relationships and edge cases, with deterministic seeds and named scenarios...

Upgrade to view official answer

Related Questions