Systematic Debugging
Find the root cause before touching code
About
A four-phase debugging method (root cause, pattern analysis, single hypothesis, fix with a failing test) that stops the agent from stacking guess fixes. After three failed fixes it tells the agent to question the architecture and talk to you.
Install
Works with Claude Code, Antigravity, Codex, Cursor, Devin CLI, Factory Droid, Gemini CLI, GitHub Copilot CLI, Grok Build CLI, Kimi Code, OpenCode, Pi, Qwen Code, Hermes Agent, Muse
Trust score
70Scored from what the skill's own files let it do. How it works
- Bundles scripts
- Runs shell commands
- Writes files in your project
- No network access
- No credentials needed
Tells the agent to add diagnostic logging, reproduce bugs and write a failing test; a supporting doc points to the bundled find-polluter.sh bisection script. Everything runs locally with no credentials or network.
Reviewed from source on Oct 15, 2026. Read the files yourself before installing.