Systematic Debugging

Forces you to find root cause before attempting any fix, no exceptions. Use it for test failures, bugs, unexpected behavior, anything broken. The core discipline is completing Phase 1 investigation (read errors, reproduce consistently, gather evidence across system boundaries, trace data flow) before even thinking about fixes. It catches the classic thrashing pattern where you're on your third attempted fix without understanding the problem. Includes a hard stop rule: if three fixes failed, question the architecture instead of trying fix number four. The "human partner signals" section is brutally honest about when you're guessing instead of investigating.

skills.sh installs: 1.3K
GitHub stars: 43.1K
last updated: Jul 14, 2026

View on GitHub

Install

$ npx -y skills add sickn33/antigravity-awesome-skills --skill systematic-debugging --agent claude-code

Installs into .claude/skills of the current project.

Files

Systematic Debugging

Overview

Random fixes waste time and create new bugs. Quick patches mask underlying issues.

Core principle: Investigate before guessing and separate a verified repair from a temporary mitigation. During an incident, an authorized rollback or containment action may restore service while root-cause work continues.

Violating the letter of this process is violating the spirit of debugging.

The Iron Law

NO FIXES WITHOUT ROOT CAUSE INVESTIGATION FIRST

Start with the observable failure and the smallest useful evidence check. Label any emergency mitigation explicitly and preserve evidence for the later root-cause investigation.