What it does
Designs production-safe synthetic checks, SLIs/SLOs, and progressive-delivery validation (canary/A-B/flags), evaluates production health, and records a promote/hold/roll-back verdict.
When to use it
When verifying quality in the live environment that pre-release testing cannot fully reach; post-deploy.
Prerequisites
qa.config.yml; live environment access; paths.*; tooling.* (e2e/api); gates (SLO thresholds); risk_areas; environments; team (approvers).
Output
ISO/IEC/IEEE 29119-3 Production Verification Report -> paths.reports_dir/production-verification-<scope>.md, plus synthetic-check scripts under paths.tests_dir.
Mechanics
How it works
- Resolve scope to the risk_areas.critical journeys (or the named feature/journey)
- Define synthetic checks / post-deploy smoke per journey against production-safe endpoints
- Define SLIs and SLOs, deriving thresholds from gates and verifying telemetry exposes each signal
- Validate progressive delivery: canary promotion/rollback criteria, A/B guardrails, feature-flag kill-switch
- Evaluate production health (healthy/degraded/roll back) and record a promote/hold/roll-back decision
- Preserve the risk-area -> check -> SLI/SLO -> result traceability and state production residual risk
Why it works
The theory behind it
Testing in production / continuous testing (Quality in DevOps). Extends coverage to real conditions but explicitly complements, never replaces, shift-left testing (Principle 3). Done safely: read-only, synthetic data, no load/destructive tests against production.
Quality in DevOps (Specialist) · ISO/IEC/IEEE 29119-3 Production Verification Report
Example
See it in use
> /qa:shift-right "checkout"✓ Correct when It defines synthetic monitors, observability checks, and canary/feature-flag validation for production.
Related commands