Automates the triage, analysis, and resolution of complex software errors using AI-driven observability and hypothesis-based debugging.
This skill provides a high-level framework for Claude to handle sophisticated debugging tasks across local and production environments. It guides the AI through a structured workflow encompassing error triage, observability data collection from platforms like Sentry and Datadog, and hypothesis generation with probability scoring. It is particularly useful for diagnosing distributed systems, identifying intermittent race conditions, and performing deep root cause analysis that goes beyond simple stack trace inspection to suggest production-safe fixes and prevention strategies.
주요 기능
01Observability integration for Sentry, Datadog, and APM metrics
02Production-safe instrumentation and dynamic logging strategies
0336,229 GitHub stars
04Fix generation with risk assessment and regression test creation
05Automated root cause analysis and execution path reconstruction
06AI-powered triage with ranked hypothesis generation
사용 사례
01Analyzing complex state management issues and race conditions in frontend or backend apps
02Diagnosing N+1 query patterns and database integration failures
03Troubleshooting intermittent production timeouts and performance bottlenecks
