What problem does it solve? Debugging and incident analysis often rely on guesswork, lost context, and unverified assumptions. This Skill enforces disciplined forensic investigation: every finding is graded as Confirmed, Deduced, or Hypothesized, and all progress is persisted in a structured case file that another engineer can pick up cold. ## Core Features & Use Cases - Evidence-graded case files: Findings, deductions, hypotheses, timelines, and missing-evidence gaps are recorded in a Markdown case file with path:line citations, so nothing is lost across interruptions. - Symptom-driven and exploration modes: Trace a bug backward from its symptom to root cause, or build a mental model of an unfamiliar code area with no symptom at all. - Premise challenging and refutation passes: The user's description is treated as a hypothesis to verify, and each near-confirmed theory is actively tested against refuting evidence. - Use Case: Given a ticket ID or a stack trace, the Skill opens a case file, maps the evidence perimeter across logs, version control, and source code, reasons about causality, and delivers a final conclusion with a confidence level and fix direction. ## Quick Start Ask the AI to investigate the error in the attached log file and produce an evidence-graded case file identifying the root cause.
