diagnose

Diagnose software bugs and performance regressions through a disciplined feedback loop.

2|Updated May 5, 2026
One-click install
npx skills add https://github.com/OrestesK/pi --skill diagnose-orestesk
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: diagnose
Source: https://github.com/OrestesK/pi/tree/main/packages/context-mode/skills/diagnose
Command: npx skills add https://github.com/OrestesK/pi --skill diagnose-orestesk

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill prevents unfocused debugging by providing a disciplined, evidence-driven process for reproducing bugs and performance regressions, identifying their causes, implementing fixes, and verifying that they stay fixed.

Core Features & Use Cases

  • Feedback Loop Construction: Create a fast, deterministic pass/fail signal using tests, scripts, HTTP checks, browser automation, trace replay, harnesses, fuzzing, bisection, or differential testing.
  • Structured Diagnosis: Progress through reproduction, ranked falsifiable hypotheses, targeted instrumentation, and controlled experiments.
  • Regression Prevention: Add tests at the correct seam, clean up debug artifacts, verify the original scenario, and document architectural improvements.
  • Use Case: When an export flow intermittently fails or a recent change causes a performance regression, use this Skill to reproduce the symptom, isolate the responsible change or condition, apply a targeted fix, and confirm the issue no longer occurs.

Quick Start

Use the diagnose skill to investigate the reported bug, build a reproducible feedback loop, identify the most likely causes, and implement a verified regression-tested fix.

Frequently Asked Questions about diagnose

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

FAQPage Schema
How do I debug intermittent software failures and performance regressions?

Debug intermittent failures and performance regressions by building a deterministic feedback loop using trace replays, scripts, or HTTP checks, then isolating causes through ranked falsifiable hypotheses and targeted instrumentation.

What is the best way to reproduce nondeterministic bugs reliably?

Reproduce nondeterministic bugs by constructing a fast pass/fail signal using browser automation, fuzzing, bisection, or differential testing to create a controlled environment that consistently triggers the unwanted behavior.

How do I isolate the specific change causing a regression between versions?

Isolate regressions between versions by applying bisection and differential testing to trace replays, forming falsifiable hypotheses about the responsible code change, and running controlled experiments to confirm the root cause.

Can I use hypothesis testing to diagnose CLI and HTTP issues?

Yes, you can diagnose CLI and HTTP issues by structuring your debugging around hypothesis testing, creating targeted scripts to reproduce the broken user flow, and instrumenting the code to verify or reject each suspected cause.

How do I prevent a fixed bug from causing future regressions?

Prevent future regressions by adding tests at the correct seam after isolating the bug, verifying the original scenario passes, cleaning up debug artifacts, and documenting architectural improvements to maintain code stability.