/rota-qa
Run product-level QA against the built artifact and score it.
When to use
Use on “/rota-qa”, “run QA”, “test the feature”, “validate the build”, before ship as a gate, or on the first cycle to scaffold a per-repo QA strategy. QA of the built product, not the diff.
Reach for another skill when
- you want a read of the diff and commits, not a run of the product:
/rota-review - nothing is built yet:
/rota-work - QA found a failure and you need to fix it:
/rota-debug
Infra preflight: dev server on :3000 reachable. Executable checks: 4 run, 4 met their pass criteria. Audit pass: no blockers. Verdict: PASS (recorded as a qa verdict; artifacts under .rota/qa-runs/<timestamp>/)
- artifacts under .rota/qa-runs/
/ - proof rows for the items on the branch
- a qa verdict that /rota-ship routes on
Flags and modes
first-run- probe surfaces and propose a strategy in .rota/qa/
.md run [<target>]- execute the strategy and return PASS, CONCERNS, FAIL or INFRA-FAIL
restructure- audit strategy files, retire dead ones, fix broken commands
uat <ID>- walk one item’s acceptance criteria with you
Details
Product-level QA: runs the per-target strategy declared in .rota/qa/<target>.md and emits a scored verdict. It does not read commits or the diff (that is /rota-review’s job). It runs the built artifact with whatever the strategy declares (Playwright, smoke scripts, contract tests, Lighthouse, axe, ZAP), grouped into executable and audit pillars. Three modes: first-run (probe surfaces, propose strategy), run (execute, score, verdict), restructure (audit strategy files). Returns PASS / CONCERNS / FAIL, or INFRA-FAIL when required infra is missing. /rota-ship invokes it when ship.qa: true; the verdict is routed by qa.gate ("advisory" reports only, "blocking" halts on FAIL). See product QA for the full flow.