Instrument Expo apps with expo-observe and query performance metrics via the EAS CLI.
npx skills add https://github.com/CYRUS-pinto/pi-bot --skill eas-observe-cyrus-pinto
Please help me install this Agent Skill. Skill: eas-observe Source: https://github.com/CYRUS-pinto/pi-bot/tree/main/.agents/skills/eas-observe Command: npx skills add https://github.com/CYRUS-pinto/pi-bot --skill eas-observe-cyrus-pinto
expo-observe to an Expo app with the correct SDK-specific API (AppMetricsRoot on SDK 55, ObserveRoot and useObserve on SDK 56+), plus per-route navigation metrics via Expo Router or React Navigation integrations. - CLI querying: Run the six eas observe:* commands (metrics-summary, metrics, routes, events, session, versions) with correct flags, aliases, pagination, and JSON output shapes. - Metrics interpretation: Read TTI frame-rate, device, and network params to distinguish slow-but-smooth launches from main-thread contention, hard freezes, or throttled devices. - Use Case: After users report slow startup, run eas observe:metrics tti --sort slowest --days 7, pull a sessionId from the JSON output, and replay the full session timeline with eas observe:session to find the blocking request. ## Quick Start Use the eas-observe skill to add expo-observe instrumentation to my Expo Router app and show me the slowest TTI samples from the last 7 days.High-intent search queries and answers about installing and using this skill.
Install expo-observe with npx expo install, wrap your root layout with AppMetricsRoot.wrap (SDK 55) or ObserveRoot.wrap (SDK 56+), and call markInteractive() once the screen is usable. You need an EAS project and a development or production build, since it does not work in Expo Go.
Use the six eas observe:* commands: metrics-summary for per-version aggregates, metrics for individual samples, routes for per-route navigation stats, events for custom logEvent data, session for a full session timeline, and versions for the app version hierarchy. All support --days, --platform, and --json.
No. expo-observe is a native library, so it requires a development build or a production build created with eas build. Metrics collected from debug builds are also not dispatched unless you set configure({ dispatchInDebug: true }).
TTR (time to first render) is measured automatically when the root is wrapped with the Observe HOC, targeting under 2 seconds. TTI (time to interactive) is not automatic; you must call markInteractive() once the screen is genuinely usable, targeting under 3 seconds including cold launch.
EAS Observe is a paid EAS service, and the server rejects queries the account's plan does not include with EAS_OBSERVE_PLAN_UPGRADE_REQUIRED or EAS_OBSERVE_FEATURE_NOT_AVAILABLE_IN_FREE_TIER. The free plan allows up to 10,000 monthly active users with limited features.
No. EAS Observe records errors as non-fatal exception events via a global handler, ObserveErrorBoundary, and Observe.reportError, but it has no crash reporting. Use Sentry or BugSnag alongside it for crash reporting.
bmiit145Instrument Expo apps with expo-observe and query performance metrics via the EAS CLI.
BEIRUXOptimize React and Next.js data fetching and bundle size in code reviews.
intelli-verse-xCapture native crashes, uncaught exceptions, and ANR events for IntelliVerseX SDK games.
expoAnalyze Expo app performance metrics with EAS Observe.