Support diagnostic evidence
Compare reproducible conditions, incomplete trials, execution evidence and usable workarounds.
Reproduce a customer failure on a stated test baseline
Keep version, input and action together so a passing different task does not contradict the report.
Change one condition when comparing diagnostic trials
Avoid attributing a result to one setting when the version and input also changed.
Record an inconclusive diagnostic test without calling it a pass
Separate failed setup, unobserved results and a completed test of the reported condition.
Keep logs and screenshots tied to the same reported execution
Avoid constructing a diagnosis from observations of different runs.
Record a usable workaround while the underlying defect stays open
Track the customer’s restored task separately from permanent repair work.
Check whether a passing support trial matches the customer environment
Compare the relevant configuration before treating a local pass as customer recovery.