What problem does it solve?
Production incidents like JVM memory leaks, Kubernetes pod crashes, slow database queries, and distributed service failures are high-stakes, time-sensitive, and often require deep platform-specific expertise to diagnose without introducing further downtime or worsening the issue.
Core Features & Use Cases
- Multi-platform debugging tooling: Step-by-step guidance for JVM diagnostics (thread dumps, heap dumps, Java Flight Recorder, async-profiler, Arthas), Python debugging (py-spy, faulthandler, memray), Kubernetes incident response (pod logs, crash analysis, ephemeral debug containers), network traffic capture, and PostgreSQL/MySQL query optimization.
- Structured incident investigation workflow: Hypothesis-driven debugging methodology, pre-change checklists, and pattern recognition for common production issues including sudden latency spikes, memory leaks, cascading failures, and CrashLoopBackOff loops.
- Real-world use case: If your e-commerce order service is experiencing intermittent OOM crashes, use this skill to capture and analyze heap dumps to identify unbounded static caches, implement a fix, and verify resolution without unplanned downtime.
Quick Start
Use the debugging-production skill to investigate the root cause of the 10x latency spike affecting your payment processing service, including reviewing recent deployments, analyzing distributed traces, and checking database query performance.
