code-debug

Reproduce, isolate, identify root causes, and verify software fixes.

7|1|Updated Mar 19, 2026
One-click install
npx skills add https://github.com/Shiien/Self-Evolved-Research-Framework --skill code-debug-shiien
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: code-debug
Source: https://github.com/Shiien/Self-Evolved-Research-Framework/tree/main/skills/code-debug
Command: npx skills add https://github.com/Shiien/Self-Evolved-Research-Framework --skill code-debug-shiien

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Systematic debugging workflow for reproducing, isolating, identifying root causes, and verifying fixes when tests fail or behavior is unexpected. Enforces the rule that three failed fix attempts in a row means the architecture is wrong, not the implementation.

Core Features & Use Cases

  • Phase 1 — Reproduce: Capture exact error with command, stdout, stderr, and environment state to enable deterministic reproduction.
  • Phase 2 — Isolate: Use binary search, git diffs, and targeted logging to locate the faulty component and change.
  • Phase 3 — Root Cause: Identify the actual cause, with guardrails to halt after three failed fix attempts if architectural issues are suspected.
  • Phase 4 — Verify Fix: Re-run the failing case and the full suite, then commit the fix with proper messaging and cleanup.

Quick Start

Describe the failure, then run the four-phase workflow to reproduce, isolate, identify the root cause, and verify the fix.

Frequently Asked Questions about code-debug

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
What is the best way to debug a failing test or unexpected software behavior systematically?

To debug systematically, reproduce the exact error with command, stdout, stderr, and environment state. This deterministic reproduction data enables reliable debugging by ensuring you can consistently replicate the software failure before isolating the faulty component.

How do I isolate a faulty component using git diff and binary search?

To isolate a faulty component, use binary search, git diffs, and targeted logging. This process locates the specific code change causing the failure by narrowing down version history and identifying the exact faulty component responsible for the unexpected behavior.

When should I stop trying to fix a bug and reconsider the architecture?

You should stop trying to fix a bug after three failed fix attempts in a row. This guardrail indicates the architecture is wrong, not the implementation, halting root cause analysis to prevent wasted effort on fundamentally flawed structural designs.

How do I verify a bug fix before committing changes to version history?

To verify a bug fix, re-run the initially failing test case and the full test suite. After confirming the fix works, commit the changes with proper messaging and cleanup to ensure no regressions were introduced during the debugging workflow.

Can I use this structured debugging workflow for CI-driven development projects?

Yes, this structured debugging workflow applies to development projects of any size, covering standalone code and CI-driven workflows. It requires deterministic reproduction data and access to version history to execute the four-phase process effectively.