Operator-owned crash telemetry for mobile games.
Capture every crash with full game context. Keep every byte on infrastructure you control — from the device, to the edge, to your on-prem triage dashboard.
Why teams choose it
Four things a generic crash reporter can’t give you.
Your data, your servers
Crash bundles travel only through infrastructure you own: a private edge on your VPS, an on-prem store, your dashboard. No third-party SDK, no vendor cloud, no CDN in the loop.
No network permissions in your game
The game process never touches the network. Transport lives in a separately signed companion app — so your store-listing privacy disclosures shrink to what players actually expect.
Game-aware, not just stack traces
Pre-allocated ring buffers record GPU state, renderer activity, and gameplay breadcrumbs — the context platform crash logs can’t see. Diagnose the bug, not just the stack.
Loss-proof by design
The crash path is async-signal-safe: no heap, no locks, no allocations. A bundle lands on disk before anything else. Uploader offline, edge down, symbols missing — every stage degrades gracefully.
How it works
One pipeline. Four stages. Each shippable on its own.
On the device
A tiny crash subsystem linked into your game writes a crash bundle to disk the instant a signal fires — then hands control back to the platform so the Play Console tombstone path still runs.
Companion uploader
A separately signed Android app accepts bundles over
a signature-checked IPC channel, stages them in its own
store, and schedules upload under Android’s work manager.
The game itself holds no INTERNET permission.
Edge buffer on your VPS
A small Go service receives, HMAC-authenticates, and rate-limits incoming bundles into a disk-backed queue — no database, no dependencies, no persistent state. Purely a transient hand-off between public internet and your private network.
On-prem processing
Over a WireGuard tunnel, your on-prem host drains the edge, symbolicates frames against your build archive, and presents grouped crashes attributed to releases. The canonical crash record lives inside your walls.
What you get
Outcomes, not integrations.
- Debuggable crashes with full GPU and gameplay context
- Shorter Play Store privacy disclosures
- GDPR-aligned data path, end-to-end
- No vendor lock-in — open formats, your infrastructure
- Symbolicated stacks resolved against your build archive
- Zero-downtime HMAC key rotation
- Partial-deployment friendly — ship stages as they land
- Platform tombstones and
ApplicationExitInfoingested as first-class inputs