What problem does it solve? Repository maintainers struggle to spot deteriorating health trends—such as rising PR latency or growing issue backlogs—hidden inside raw metrics data, and often lack a systematic way to investigate root causes before proposing workflow changes. ## Core Features & Use Cases - Time-Series Trend Analysis: Loads metrics-timeseries.csv to detect anomalies like increasing PR latency, growing open issues, or abnormal Actions and Gemini usage costs. - Hypothesis-Driven Root Cause Investigation: Develops competing hypotheses, gathers evidence via the gh CLI and GraphQL, and selects the explanation best supported by data. - Workload-Aware Recommendations: Assesses maintainer capacity and identifies whether bottlenecks sit with authors, maintainers, or CI infrastructure before proposing interventions. - Use Case: A maintainer notices review turnaround slowing over several weeks; this Skill analyzes the metrics history, tests hypotheses about reviewer overload versus CI delays, and proposes a systemic fix such as automated routing instead of more pings. ## Quick Start Analyze the repository metrics history to find the most significant deteriorating trend, investigate its root cause, and report your findings.