Essential session storage is required to sign in. Optional analytics remain off unless you allow them.
Act as QA lead for a feature release end-to-end. Deliver: (1) a TEST PLAN (scope, approach across levels, risks, entry/exit criteria); (2) test cases mixing functional and at least one non-functional dimension (e.g. a load test with k6/JMeter, or an accessibility pass); (3) a defect workflow with well-formed bug reports that correctly separate SEVERITY from PRIORITY; (4) an automated regression suite in CI plus a documented strategy for handling FLAKY tests (root-cause categories — timing/order/shared-state — quarantine, fix; not blind re-runs); and (5) a TEST SUMMARY REPORT with honest metrics (pass rate, defect density, open defects by severity, residual risk) and a clear go/no-go release recommendation. Avoid vanity metrics.
Coherent test plan across levels with risk-based prioritization and clear entry/exit criteria; functional + a real non-functional dimension covered.
Bug reports reproducible and correctly separate severity (impact) from priority (urgency); flaky-test strategy roots out causes rather than re-running blindly.
Test summary with honest, decision-useful metrics (pass rate, defect density, open-by-severity, residual risk) and a justified go/no-go; no vanity metrics.
Regression suite wired into CI as a quality gate; appropriate pyramid balance.
Use this README structure to make the work reviewable.
DevPath fetches a read-only, size-capped snapshot. Submitted code is never executed.
# A QA strategy: plan, defects, flaky tests & metrics ## Goal Act as QA lead for a feature release end-to-end. Deliver: (1) a TEST PLAN (scope, approach across levels, risks, entry/exit criteria); (2) test cases mixing functional and at least one non-functional dimension (e.g. a load test with k6/JMeter, or an accessibility pass); (3) a defect workflow with well-formed bug reports that correctly separate SEVERITY from PRIORITY; (4) an automated regression suite in CI plus a documented strategy for handling FLAKY tests (root-cause categories — timing/order/shared-state — quarantine, fix; not blind re-runs); and (5) a TEST SUMMARY REPORT with honest metrics (pass rate, defect density, open defects by severity, residual risk) and a clear go/no-go release recommendation. Avoid vanity metrics. ## Acceptance criteria - [ ] Plan & strategy: Coherent test plan across levels with risk-based prioritization and clear entry/exit criteria; functional + a real non-functional dimension covered. - [ ] Defects & flakiness: Bug reports reproducible and correctly separate severity (impact) from priority (urgency); flaky-test strategy roots out causes rather than re-running blindly. - [ ] Metrics & reporting: Test summary with honest, decision-useful metrics (pass rate, defect density, open-by-severity, residual risk) and a justified go/no-go; no vanity metrics. - [ ] Automation in CI: Regression suite wired into CI as a quality gate; appropriate pyramid balance. ## Verification - [ ] Document installation and run commands. - [ ] Record automated checks and their exact commands. - [ ] Add representative output, screenshots, or a short demo where useful. - [ ] Confirm failure paths and known constraints. ## Decision log ### Decision title - Context: - Choice: - Alternatives considered: - Tradeoffs: ## Evidence - Link each rubric criterion to the file, test, or artifact that demonstrates it. ## Known limitations - List what is intentionally out of scope and what should be improved next.