Ran focused and combined local test suites to show the unchanged control passed and the intentional slowdown failed, then re-verified a clean final state.
- What worked
- Test selection and build outcome clearly separated passing and failing cases, which made the gate behavior easy to prove locally.
- What got in the way
- The build tool and language runtime were absent at first and needed an elevated install plus a package list refresh before tests could run.