Menu
Prefer fastest Maestro wait timeouts via shared TIMEOUT_* ladder; use when writing or editing apps/mobile/e2e YAML flows.
Instalar com Codex ou Claude Copie este prompt, cole no Codex, Claude ou outro assistente e deixe que ele revise a página da skill e instale para você.
Baseado na classificação ocupacional SOC
Just-in-time phase workflow for the Podverse mobile master plan. Use when the operator asks what to work on next in the mobile master plan, to plan or detail the next mobile phase, to create deferred detail docs and COPY-PASTA sets, or to implement a mobile master-plan phase.
Cross-platform theme AND visual parity for apps/mobile — same theme IDs and design tokens as web, RN ThemeProvider, uit pref, and web-referenced screen layouts. Use when adding mobile styling, ThemeProvider, theme selector, design token imports, or building any mobile screen that has a web counterpart.
FOSS vs playstore Android/iOS product flavors — FCM vs UnifiedPush, non-FOSS dependency register, defer full implementation to Track 20.
Expo SDK 52 as a standalone install under apps/mobile — own lockfile, file: shared packages, Metro. Use when mobile npm install, prebuild, pod install, mobile:dev fails, or when telling the operator how to run expo install (never bare root npx expo).
abcmemory checklist for mobile work — when to update .cursor/ rules/skills vs .llm/plans; abcremember vocabulary; links mobile-master-plan-phasing.
Offline-first mobile data layer — SQLite/Drizzle repositories, background sync, storage boundaries. Use when adding or changing apps/mobile data fetching, local persistence, queue/history/add-by-RSS storage, or sync.
| name | mobile-maestro-timeouts |
| description | Prefer fastest Maestro wait timeouts via shared TIMEOUT_* ladder; use when writing or editing apps/mobile/e2e YAML flows. |
Keep Maestro waits short. Fail fast; bump only when a run proves a tier is too aggressive.
A failing flow that waits the full timeout on a doomed assertion wastes operator time on every run. Minimize wall-clock-to-failure:
TIMEOUT_SLOW (8s), not
TIMEOUT_SLOWEST. When it fails, you eat 8s, not 20s+.extendedWaitUntil on the decisive testID per phase.
Avoid asserting several intermediate elements that each carry their own timeout.tab-home after login). If it never appears, the single wait fails once instead of many.launchApp / connect steps via e2e/shared/ and don't
relaunch mid-flow unless the test requires it.apps/mobile/e2e/shared/timeouts.env)| Name | ms | Typical use |
|---|---|---|
TIMEOUT_FASTEST | 2000 | Immediate UI swap already on-screen |
TIMEOUT_FASTER | 3000 | Quick local navigation |
TIMEOUT_FAST | 5000 | Default after connect / most asserts (connect-dev-client) |
TIMEOUT_SLOW | 8000 | Network-backed login / first authenticated shell |
TIMEOUT_SLOWER | 12000 | Health polls / slower API readiness |
TIMEOUT_SLOWEST | 20000 | Last resort only |
scripts/mobile/e2e-test.sh sources that env file and passes -e TIMEOUT_*=… into Maestro.
In YAML use ${TIMEOUT_FAST} (etc.), never invent ad-hoc 15000 / 60000 literals.
TIMEOUT_FAST over TIMEOUT_SLOW.e2e/shared/ so timeouts stay consistent.waitForAnimationToEnd without need). Prefer extendedWaitUntil on a testID.auth-login, auth-logout, tab-switch-playback, api-health need Mobile E2E API
(npm run mobile:e2e:api on :4230) leave-running. Restarting Metro/mobile:dev:e2e does
not require restarting the API if it is already healthy. e2e-test.sh exits early if :4230
is down for those flow names.