What problem does it solve? Incident investigations often suffer from speculation, vague claims, and premature conclusions. This Skill enforces disciplined SRE practices so findings are backed by evidence, confidence levels are stated honestly, and reports follow a consistent structure. ## Core Features & Use Cases - Confidence Calibration: Classify findings as high confidence, likely, possible, or uncertain, and clearly separate observed facts from inferred hypotheses. - Evidence-Based Reasoning: Require every claim to cite specific data such as timestamps, metric values, and error messages, and build chronological incident timelines. - Falsification and Thoroughness: Actively seek contradicting evidence, consider alternative causes, and push past surface symptoms to actionable root causes. - Use Case: During a Kubernetes outage, use these principles to produce a report stating the root cause (e.g., OOMKilled pods from a memory leak in a specific commit), the supporting metrics, ruled-out hypotheses, and concrete remediation steps. ## Quick Start Apply the SRE investigation principles to analyze the current incident and produce a structured root cause report with evidence and confidence levels.
