What problem does it solve? Debugging and code exploration often rely on guesswork and lose context across sessions. This Skill reconstructs what happened from available evidence, grades every finding as Confirmed, Deduced, or Hypothesized, and persists everything in a structured case file another engineer can pick up cold. ## Core Features & Use Cases - Evidence-graded investigation: Anchors on one Confirmed piece of evidence (a stronghold), traces causality backward from symptoms or outputs, and records hypotheses with explicit confirm/refute criteria that are never deleted. - Persistent case files: Initializes a Markdown case file from a template covering evidence inventory, timeline, backlog, source code trace, and a confidence-rated conclusion, with follow-up blocks for multi-session work. - Dual-mode calibration: Handles symptom-driven defect chasing (tickets, logs, stack traces) and symptom-free code area exploration with the same disciplined workflow. - Use Case: Given a ticket ID and a log file, the Skill maps the evidence perimeter, traces the error to its origin with path:line citations, and hands off a final report with fix direction and reproduction plan. ## Quick Start Ask the AI to investigate the bug described in ticket PROJ-1234 using the attached error log and produce an evidence-graded case file.
