Rejourney: open-source session replay and crash reporting for Flutter

Rejourney is an open-source platform for session replay, product analytics and crash/ANR reporting, with a Flutter SDK that is more carefully built than most. 277★, last pushed 2026-08-28.

What is Rejourney?

Most teams debug conversion problems with three disconnected tools: an analytics dashboard that says the funnel dropped, a crash reporter that says something threw, and a session replay product that shows one user’s afternoon. Correlating them is manual work.

Rejourney’s premise is to keep them in one place. You instrument a handful of critical conversion events — checkout_started, purchase_completed — and it records the surrounding journey, the interaction data (touches, scrolls, pans, rage taps), and the technical evidence (API latency and status codes, errors, crash traces, ANRs). Similar sessions get grouped into cohorts, and when a cohort trends badly around a conversion event, the replays and evidence surface as a ranked issue.

Beside the replay workbench sit journey maps, interaction heatmaps, endpoint views, device and geographic cohorts, and project-level analytics for version adoption, retention and — where a revenue source is connected — revenue impact.

Why the Flutter SDK is worth a look

Plenty of analytics products bolt a thin Dart wrapper onto a native SDK. Rejourney splits it the other way round: capture runs in native code, while the public API, navigation integration, privacy masks, error hooks and HTTP client are idiomatic Flutter.

That shows up in the details:

• RejourneyNavigatorObserver plugs into navigatorObservers and handles push, pop, replace and remove, suppressing duplicate screen names. With a declarative Router, you call Rejourney.trackScreen('checkout') from its route-change callback instead.
• RejourneyMask hides a Flutter-rendered subtree from captured frames without changing what the user sees, and tracks layout changes, scrolling and disposal.
• RejourneyHttpClient wraps package:http to record method, URL, status, timing, content type and byte sizes — not bodies. For Dio, gRPC or GraphQL you call Rejourney.logNetworkRequest(...) from an interceptor.
• RejourneyErrorCapture.install() goes in before runApp.

The Android capture path deserves special mention. Some renderer and device combinations report a successful PixelCopy while handing back an entirely black bitmap. Rather than telling you to switch render modes, the SDK detects the false success and captures Flutter’s retained layer tree at reduced resolution instead, leaving your live FlutterSurfaceView alone. getSdkMetrics().lastCaptureSource tells you which path is active. That is the kind of problem you only fix after shipping to real devices.

Getting started

flutter pub add rejourney

Minimums: Flutter 3.22, Dart 3.3, iOS 15.1, Android API 24. On an existing iOS app you may need cd ios && pod install.

import 'package:rejourney/rejourney.dart';

await Rejourney.init('pk_live_your_public_key');
await Rejourney.start();

init never starts capture — that separation exists so you can wait for consent:

if (await consentStore.canRecord()) {
  final result = await Rejourney.start();
  debugPrint('session=${result.sessionId} replay=${!result.telemetryOnly}');
}

Call Rejourney.stop() if consent is revoked. Then add the navigator observer, wrap sensitive inputs in RejourneyMask, and log your conversion events:

await Rejourney.logEvent('purchase_completed', <String, Object?>{
  'transactionId': order.id,
  'amount': order.total,
  'currency': 'USD',
});

When should you use Rejourney?

• you want replay, analytics and crash reporting correlated rather than in three tabs
• ANRs and native crashes matter to you as much as Dart exceptions
• data residency or cost rules you out of a hosted analytics vendor
• you need per-widget masking rather than a global “hide all text” switch

Where it falls short

The licence is dual, and the split matters. Client SDKs, examples and docs are Apache 2.0 — safe to ship. The backend and dashboard are SSPL 1.0: fine for internal use, but if you offer Rejourney as a service you must publish your service source. Anyone treating “open source” as a synonym for “no obligations” should read LICENSE before building a product on it.

Self-hosting a replay backend is not a weekend job either. You are running ingestion, storage and a dashboard for video-like data, and the storage bill grows with your traffic, not your team.

The project is also honest that its issue reports are heuristic — a reported pattern is “a starting point for investigation, not proof of causality”. Take that seriously. A cohort correlation plus a suggested fix is a hypothesis, and shipping it without checking your authoritative product or payment state is how you fix the wrong thing confidently.

Finally: session replay on a real product is a privacy decision before it is a technical one. Consent, masking and replay QA on the exact login, checkout and health flows you ship are not optional extras here.

Alternatives worth comparing

• Sentry — mature crash and performance monitoring, no product analytics or journey view
• PostHog — the closest open-source analogue, stronger on web, thinner on native mobile capture
• Add AI to your Flutter app with appwrite and Build better Flutter UI with serverpod — self-hosted backends you might already run

Frequently asked questions

How do I add Rejourney to a Flutter app?

flutter pub add rejourney, then await Rejourney.init('pk_live_...') at startup and await Rejourney.start() only after your consent decision. init deliberately never starts capture. Minimums are Flutter 3.22, Dart 3.3, iOS 15.1 and Android API 24.

Is Rejourney really open source?

Partly, and it is upfront about it. The Flutter API, platform bridges, native core, examples and documentation are Apache 2.0. The backend and dashboard are SSPL 1.0 — free to run internally, but offering it as a service obliges you to publish your service source.

How does it handle sensitive fields?

Wrap the subtree in RejourneyMask, which hides a region in captured frames without changing what the user sees. That sits on top of native secure-field detection and project-level text and media privacy rules, and remote policy can only make capture more restrictive, never less.

Does session replay work reliably on Android?

Mostly. Some renderer and device combinations report a successful PixelCopy while returning a black bitmap; from 0.2.1 the SDK detects that and falls back to capturing Flutter’s retained layer tree at reduced resolution. Check getSdkMetrics().lastCaptureSource when validating a device.

Resources & links

• GitHub: rejourneyco/rejourney
• pub.dev: rejourney
• Website: rejourney.co

---

Part of FlutterCook — hands-on guides to the best open-source Flutter libraries, UI kits, and apps. Explore the live GitHub trends or browse YouTube guides.

---

Originally published on FlutterCook. Read the latest version there — that copy is the one kept up to date.