Skip to content
OpenAgentSkill
Browsing is open · Catalog captured 2026-11-21 · Automatic updates and account services are paused.
Design & UI

design-review-workflow

Enforces one structured end-of-build design/QA pass — never per-task — on the release build: a screenshot sweep matrix (every screen × light/dark × LTR+RTL × largest text scale × reduce-motion) plus an on-device pass on real cheap target hardware, findings graded BLOCKER/FIX/NOTE with every accessibility-floor violation a mandatory BLOCKER, exactly one scoped fix round, and a dated sign-off artifact that gates release. Use when a feature's build tasks are done and it needs its design review, tagging or preparing a release, verifying a build on a physical device, running a pre-ship visual/QA sweep, or asking what design and QA cannot be tested automatically.

zakariaf · Static checks on file

Installation

Read the source and dependencies before running a command. Copying does not install or execute anything.

npx skills add zakariaf/Flutter-Skills --skill design-review-workflow

About this skill


name: design-review-workflow description: Enforces one structured end-of-build design/QA pass — never per-task — on the release build: a screenshot sweep matrix (every screen × light/dark × LTR+RTL × largest text scale × reduce-motion) plus an on-device pass on real cheap target hardware, findings graded BLOCKER/FIX/NOTE with every accessibility-floor violation a mandatory BLOCKER, exactly one scoped fix round, and a dated sign-off artifact that gates release. Use when a feature's build tasks are done and it needs its design review, tagging or preparing a release, verifying a build on a physical device, running a pre-ship visual/QA sweep, or asking what design and QA cannot be tested automatically.

Design review workflow

One structured design/QA pass per app or feature, at the very end, on the release build — never a per-task visual critique. During the build, development speed wins and no task is blocked on aesthetics; when the last build task is done and CI is green, exactly one pass runs: a screenshot sweep + an on-device pass → graded findings → one scoped fix round → a dated sign-off that gates release.

Non-negotiable rules

  1. Once per app/feature, at the end — never per-task. Blocking a build task on a visual critique, or demanding review screenshots inside a task's acceptance criteria, is the failure this skill prevents. If something looks broken mid-build, file a note and keep moving; review it in the pass.
  2. Trigger = last build task done AND all CI gates + tests green. Green gates are a precondition, not part of the review — this pass judges only what the eye and ear catch. Never re-litigate what CI already proves (determinism, lint, l10n key parity, layer gates).
  3. Review the release build, not a debug build. A debug banner, debug-mode jank, or an un-overridden status bar wastes the sweep. Standardize the status bar (fixed clock, full battery) so shots differ only where the UI differs.
  4. Sweep the full matrix — every screen the app can show. Each screen × {light, dark} × {LTR, RTL} × largest text scale × reduce-motion. Dynamic screens get shot in each meaningful state (empty, mid, full/success). One file per matrix cell, machine-sortable names, all artifacts under one review folder inside the write-set.
  5. Every accessibility-floor violation is a BLOCKER — however beautiful the screen. Contrast, tap-target size, text-scale reflow, color-never-alone, RTL correctness, and reduce-motion safety are the shared floor. A floor miss always grades BLOCKER; aesthetics never buy it back.
  6. Grade every finding BLOCKER / FIX / NOTE. BLOCKER = floor violation or a shipping-stopper (must fix now); FIX = clearly below bar, fixable this round; NOTE = recorded, deferred to the backlog. Consolidate into one deduped table before fixing.
  7. Exactly one fix round. Fix BLOCKERs + FIXes as a single scoped unit, re-run gates/tests, re-shoot only affected cells (overwrite — the folder stays one truth), verify each finding against its new shot. Open no new critique during verification — new observations become NOTEs. A surviving BLOCKER means no sign-off and an escalation, not a second round.
  8. Do destructive on-device steps last. Force-stop, previous-release migration, wipe/reinstall, and a deliberate crash each destroy the state before them. Ordering them last means the audio/reader/scale passes aren't redone.
  9. Ship a dated sign-off artifact that gates release. Date, reviewer, commit sha, build flavor, the matrix inventory, the findings table with resolutions, and a verdict line (SIGNED OFF, or NOT signed off + the escalated blocker). It is a tracked file, not a loose note, and release tasks depend on it.

The screenshot sweep matrix

Walk every screen in screen order, once per matrix cell. Any screenshot mechanism works — simulator/emulator capture (xcrun simctl io booted screenshot …, adb exec-out screencap), an integration-test golden harness, or a device screen-grab.

Multiplying axes (one still per cell):
  screen    every screen the app can reach + each meaningful state
  theme     light, dark        (flip the OS appearance; leave in-app on "System")
  direction LTR, RTL           (an RTL locale via the in-app locale picker)

Applied to every still (not a separate axis):
  textScale largest supported  (OS font-size max + bold text) — every still is shot under it

Video-only (never a still):
  motion    reduce-motion off, on   (record motion moments as video)
  • Name one file per cell, sortable: NN--<screen>[-<state>]--<theme>--<dir>.png. The cell axes are screen×state × theme × direction, so the name needs no textScale or motion token — textScale is applied to every still, and motion evidence is a short video per moment (skippability and the reduce-motion end-state are only judgeable on video).
  • Dark theme must be a designed palette, not an inverted afterthought.
  • Largest text scale must reflow without truncation or clipping on the smallest supported device — the case a clamp hides. Never clamp the text scaler to keep a layout tidy; that is a floor violation (see accessibility-as-code).
  • RTL cells check: chrome mirrors, directional insets applied, no bidi garbling in mixed runs, locale-correct numerals, and no missing-glyph tofu in the chosen faces (see i18n-rtl-l10n).

Grading rubric

Critique each screen through four lenses; grade each finding.

  • Floor compliance (absolute). Contrast on the worst offender per screen in both themes, every daily-flow control at a real minimum tap size, desaturate a shot to prove no state relies on color alone, reduce-motion shows instant end-states, RTL correct. Any miss = BLOCKER.
  • Identity fidelity. Does the screen deliver the app's declared design language (its theme/tokens/components) — not a generic default? Judge delivery of the declared identity; changing the identity is a design-system amendment, out of scope here.
  • **Parity or bett
Read the complete source ↗