Faultbook από tranminh6372

Faultbook is a local incident-reconstruction workbench. It separates what was expected, what signal was observed, what assumption was held, what decision was made, and what outcome followed. The sequence stays visible as a horizontal evidence strip while the selected event can be corrected or expanded.

A second view builds a control ladder across elimination, prevention, detection, limitation, recovery, and reminders. Faultbook audits whether an incident has meaningful detection and recovery coverage, whether assumptions were recorded, and whether signals were connected to decisions. The resulting coverage score is a prompt for a stronger system, not a score for a person.

An included package incident demonstrates how an apparently reasonable release decision can expose missing controls. A self-contained HTML report can be exported for a review meeting or local archive. Faultbook contains no collaboration backend and sends no incident information anywhere.