software-quality-assurance-analysts-and-testers Post-execution QA gate for phased AI development workflows (GSD, Superpowers, any phased build). Runs AFTER `/gsd-execute-phase` completes — execute-phase is NOT part of this skill. Orchestrates, in order: (1) `/gsd-code-review` + `/gsd-code-review-fix --auto`, (2) regression gate (tsc + build + lint), (3) `/codex:review` + apply fixes, (4) regression gate again, (5) parse GSD phase artifacts (PLAN.md / SUMMARY.md) to derive UAT scenarios per requirement, (6) wire auth bypass via smoke-test-auth-bypass skill, (7) generate + run a Playwright spec file headlessly with DOM + dimension + console + network asserts, (8) DB assertions via Supabase MCP for mutation scenarios, (9) auto-fix loop + regression gate, (10) conditional `/gsd-verify-work` — only if user-facing surface changed (routes, components, middleware, server actions), skipped for pure infra / API / migration phases, (11) `/gsd-extract_learnings` to persist phase decisions. Trigger on: "smoke test", "run tests", "QA this", "verify the build", phase com
2026-04-24