| name | regression-sweep |
| description | Run a broad health check across a Nimara surface — checkout, SEO/metadata, performance, or a page-type matrix — and report findings, instead of verifying one ticket. Use this skill whenever the user asks for a regression pass, a smoke/health check, a "sweep" of an area, a performance audit (Lighthouse), an SEO/meta check, or "test everything on <page/flow>". Fans out across page types, channels, and devices, then summarises pass/fail with evidence. Do NOT use to retest a specific reported bug (use bug-retest-triage) or to design cases from a spec (use test-case-design). |
Regression Sweep
Run a broad, repeatable health check across a surface and report — breadth over depth. Use this for "is anything broken across X" rather than a single defect.
Operating principles
- Cover the matrix. Enumerate the axes for the surface (page type × channel × device × env) and sweep representatives — don't spot-check one page. See [[Coverage Maps]].
- Cheapest reliable method per signal. curl+grep for SEO/structural, local Lighthouse for perf, Playwright for interactive flows, response listeners for contracts. See [[Test Method Playbooks]].
- Right environment. Test the real storefronts (dev/stage/demo), not the
nimara.store landing page; mind channel prefixes (/ = US, /gb = GB). See [[Environments & Access Matrix]].
- Report findings, don't auto-file. Summarise pass/fail with evidence; only open tickets when the user asks, generalising to classes.
Workflow
- PICK SURFACE & AXES — e.g. SEO across {Home, PLP, PDP, CMS} × {US, GB}; or perf across {dev, stage, demo} × {Home, PLP, PDP} × {mobile, desktop}; or checkout smoke across {US, GB} × {guest, logged-in}.
- BASELINE (if any) — note prior scores/expected values to compare against.
- SWEEP — run each cell with its method; capture evidence under
qa/triage/evidence/<sweep-name>/. Independent cells parallelise well.
- REPORT — matrix of cell → result → delta-from-baseline; flag regressions and known-flaky cells ([[Known Flaky, Blocked & Backend-Only]]); record method caveats (perf fluctuates, code-inspection vs deployed build).
- HANDOFF (if asked) — turn confirmed regressions into bugs via [[Defect Taxonomy & Severity]] + [[Jira & Board 74 Operating Manual]].
References
[[Coverage Maps]] · [[Test Method Playbooks]] · [[Environments & Access Matrix]] · [[Known Flaky, Blocked & Backend-Only]]
