debug

Trace callstacks and diagnose root causes with filename:line logging.

36|3|Updated Jan 6, 2026
One-click install
npx skills add https://github.com/2lab-ai/oh-my-claude --skill debug-2lab-ai
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: debug
Source: https://github.com/2lab-ai/oh-my-claude/tree/main/plugins/stv/skills/debug
Command: npx skills add https://github.com/2lab-ai/oh-my-claude --skill debug-2lab-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Provides a disciplined, repeatable methodology for diagnosing software that behaves differently than expected by tracing execution, recording findings, and isolating root causes to prevent regressions.

Core Features & Use Cases

  • Blackbox callstack tracing: Follow the callstack step-by-step across files and services while writing exact filename:line observations to a persistent trace log.
  • Hypothesis-driven exploration: Start with a top-3 heuristic check then move to exhaustive branch exploration, ensuring no branch remains unexamined.
  • Red-Green verification cycle: Reproduce the issue with a failing test, implement a focused fix, and confirm the fix plus regression safety.
  • Use Case: Use when a production or test failure is non-obvious—record the trace, form single hypotheses per iteration, and converge to a reliable fix with instrumentation and layered guards.

Quick Start

Start at the failing entry point, follow the callstack one step at a time into each file or service, and record each filename:line and conditional outcome into the trace log while iterating on single hypotheses and writing a reproduce-first test.

Frequently Asked Questions about debug

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

FAQPage Schema
How do I find the root cause of a flaky test failure across multiple services?

Trace the callstack step-by-step across files and services to find the root cause of a flaky test, recording exact filename:line observations and conditional outcomes into a persistent trace log to isolate the issue.

What is blackbox callstack debugging and when should I use it?

Blackbox callstack debugging is a methodology for diagnosing unexpected software behavior by following the callstack and recording each executed step. Use it when a production or test failure is non-obvious and requires reproducing failures and instrumenting traces.

How do I debug an unexpected conditional branch outcome in my code?

Debug an unexpected conditional branch by starting with a top-3 heuristic check, then moving to exhaustive branch exploration to ensure no branch remains unexamined. Record each filename:line and conditional outcome into a trace log while iterating on single hypotheses.

What's the best way to reproduce a software failure and verify the fix?

The best way to reproduce a software failure and verify the fix is to use a red-green verification cycle: reproduce the issue with a failing test, implement a focused fix, and confirm the fix plus regression safety with layered guards.

How do you trace execution across services when heuristics fail to find the bug?

Trace execution across services by following the callstack one step at a time into each file or service, writing exact filename:line observations to a persistent trace log, and applying exhaustive branch exploration when heuristic checks fail to locate the root cause.

Does this debugging approach work without setting up external instrumentation dependencies?

Yes, this debugging approach works without external dependencies by using a blackbox tracing methodology. It follows the callstack step-by-step, recording each executed step and conditional outcome into a persistent trace log to locate root causes.