use-sentry
Installation
SKILL.md
Sentry
The complete enterprise Sentry observability lifecycle: autonomous stack research & zero-to-one setup, multi-pillar observability auditing (OpenTelemetry, Continuous Profiling, Crons, Metrics, Session Replay, Spotlight), and AI-accelerated production incident triage with Seer AI.
The 7 Enterprise Observability Pillars
Modern Sentry extends beyond basic error catching into a unified, full-stack observability platform:
- Error Monitoring & Exception Capture: Automated unhandled exception capture, source maps, contextual breadcrumbs, and semantic issue grouping.
- OpenTelemetry & Distributed Tracing: Native OpenTelemetry engine (
@sentry/opentelemetry), W3Ctraceparentandbaggagepropagation, and end-to-end distributed span waterfall trees (Sentry.startSpan). - Continuous Profiling: Low-overhead runtime CPU and memory profiling (
@sentry/profiling-node,profileSessionSampleRate) decoupled from individual transactions. - Cron & Uptime Monitoring: Heartbeat check-ins and missed execution alerts for background queues and scheduled tasks (
Sentry.withMonitor,Sentry.captureCheckIn). - Application Metrics: Real-time telemetry counters, distributions, gauges, and sets (
Sentry.metrics.increment,distribution,gauge,set). - Session Replay & Privacy Redaction: Video-like DOM reconstruction correlated with error traces, enforced with strict privacy shields (
maskAllText,blockAllMedia) and network allowlists. - Spotlight Local Development Overlay: Zero-overhead local development sidecar and interactive browser overlay (
@spotlightjs/spotlight,spotlight: true) for instant trace visibility.