Apply systematic debugging methodology using medical differential diagnosis principles. Trigger when AI modifies working code and anomalies occur, or when users report unexpected test results or execution failures. Use observation without preconception, fact isolation, differential diagnosis lists, deductive exclusion, experimental verification, precise fixes, and prevention mechanisms.
View on GitHubiciakky/cc-general-skills
cc-general-skills
/skills/debug/SKILL.md
January 15, 2026
Select agents to install to:
npx add-skill https://github.com/iciakky/cc-general-skills/blob/main//skills/debug/SKILL.md -a claude-code --skill debugInstallation paths:
.claude/skills/debug/# Debug
## Overview
This skill applies a systematic debugging methodology inspired by medical differential diagnosis. It provides a rigorous 7-step process for investigating and resolving bugs through observation, classification, hypothesis testing, and verification. This approach prioritizes evidence-based reasoning over assumptions, ensuring root causes are identified rather than symptoms treated.
## When to Use This Skill
Activate this skill in two primary scenarios:
**Scenario A: Post-Modification Anomalies**
When modifying a previously tested and working version, and any unexpected behavior emerges after the changes.
**Scenario B: User-Reported Issues**
When users report that test results don't meet expectations or the system fails to execute as intended.
## Debugging Workflow
Follow this 7-step systematic approach to diagnose and resolve issues.
For a detailed checklist of each step, refer to `{baseDir}/references/debugging_checklist.md`. For common bug patterns and their signatures, see `{baseDir}/references/common_patterns.md`.
### Step 1: Observe Without Preconception (Observe)
**Objective:** Collect all available evidence without jumping to conclusions.
**Process:**
- Gather all accessible clues: user reports, system logs, dashboards, error stack traces, version changes (git diff), configuration parameters (configs/args/env)
- Focus exclusively on facts and observable phenomena
- Avoid premature hypotheses or assumptions about causes
- Document all observations systematically
**Key Principle:** Observe, don't just see. At this stage, the goal is comprehensive data collection, not interpretation.
### Step 2: Classify and Isolate Facts (Classify & Isolate Facts)
**Objective:** Distinguish symptoms from root causes and narrow the problem scope.
**Process:**
**For Incremental Development (Scenario A - Post-Modification Anomalies):**
- Confirm the previous step still works (ensure issue is from new changes)
- List ALL changes since last working