diagnose

Diagnose software bugs and performance regressions with a structured debugging workflow.

46|7|Updated Oct 12, 2025

Search Tags:#performance #debugging #bug fix #instrumentation #hypothesis #reproduction

Ready to run

npx skills add https://github.com/adriannoes/awesome-vibe-coding --skill diagnose-adriannoes

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: diagnose
Source: https://github.com/adriannoes/awesome-vibe-coding/tree/main/cursor-and-claude/skills/matt-pocock/engineering/diagnose
Command: npx skills add https://github.com/adriannoes/awesome-vibe-coding --skill diagnose-adriannoes

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill provides a disciplined approach to identifying and resolving hard bugs and performance regressions in software projects, reducing debugging time and increasing reliability.

Core Features & Use Cases

Structured Debugging Process: Guides users through feedback loops, reproduction, hypothesis generation, instrumentation, fixing, and cleanup.
Application Scope: Suitable for developers and engineers troubleshooting code issues, performance issues, or regressions across various environments.
Use Case: When facing a hard bug or performance slowdown, use this Skill to systematically isolate the cause, validate hypotheses, and apply fixes efficiently.

Quick Start

Follow the steps outlined to build a feedback loop, reproduce the bug, hypothesize causes, instrument code, implement fixes, and clean up after resolving the issue.

Frequently Asked Questions about diagnose

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

FAQPage Schema

What is a systematic approach to debugging hard bugs in complex codebases?▼

A systematic debugging workflow structures investigation through feedback loops, deterministic reproduction, hypothesis generation, instrumentation, and regression testing. This methodology reduces time-to-fix and improves bug resolution accuracy for difficult software issues.

How do I reproduce a performance regression for hypothesis-driven investigation?▼

Build a deterministic test loop to reliably reproduce the performance regression. Once reproduced, generate hypotheses about the cause, instrument the code to capture execution metrics, and validate your assumptions before applying fixes.

What is the best way to isolate the cause of a performance slowdown in software?▼

The best way to isolate a performance slowdown is through hypothesis-driven investigation combined with code instrumentation. By systematically measuring specific execution paths and validating hypotheses against data, you can pinpoint the root cause efficiently.

Does structured debugging work for both functional bugs and performance issues?▼

Yes, structured debugging works for both functional bugs and performance issues across various environments. The methodology scales to troubleshoot code regressions and performance slowdowns by applying consistent feedback loops, reproduction, and instrumentation steps.

How do I clean up instrumentation code after fixing a bug?▼

Post-mortem cleanup is the final step in the debugging workflow. After resolving the issue and validating the fix through regression testing, remove all temporary instrumentation code added during the investigation to restore the codebase to its original state.