diagnose

Name: diagnose
Availability: InStock
Author: ste-bah

Diagnose hard bugs and performance regressions through a deterministic feedback loop.

7|Updated Apr 4, 2026

Search Tags:#performance #debugging #bisection #instrumentation #diagnose #regression-testing

Ready to run

npx skills add https://github.com/ste-bah/archon-cli --skill diagnose-ste-bah

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: diagnose
Source: https://github.com/ste-bah/archon-cli/tree/main/assets/skills/diagnose
Command: npx skills add https://github.com/ste-bah/archon-cli --skill diagnose-ste-bah

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you systematically diagnose hard bugs and performance regressions by building a reliable feedback loop before guessing at a fix.

Core Features & Use Cases

Reproduce the issue with the smallest possible failing case so the symptom is consistent and measurable.
Rank falsifiable hypotheses and test them one by one with targeted instrumentation instead of broad, noisy logging.
Lock in the fix by turning the minimal repro into a regression test and verifying the original issue no longer occurs.
Use Case: When a CLI command intermittently fails or a new release slows down, this Skill guides you through reproducing, isolating, and proving the root cause.

Quick Start

Use the diagnose skill to reproduce the bug, narrow it to a minimal failing case, and convert that repro into a regression test.

Frequently Asked Questions about diagnose

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

FAQPage Schema

How do I systematically debug a flaky test or intermittent CLI failure?▼

Debugging an intermittent CLI failure requires building a deterministic feedback loop by reproducing the issue with the smallest possible failing case, ensuring the symptom is consistent and measurable before guessing at a fix.

What is the best way to isolate a performance regression after a new release?▼

Isolating a performance regression requires building a deterministic feedback loop using bisection and targeted instrumentation to rank falsifiable hypotheses, proving the root cause rather than relying on broad logging.

How do you create a regression test from a minimal reproduction case?▼

Creating a regression test from a minimal reproduction case involves converting the repro into a test at the correct seam, verifying the original issue no longer occurs, and locking in the fix against future regressions.

Does targeted instrumentation work better than broad logging for finding root causes?▼

Targeted instrumentation works better than broad logging for finding root causes because it tests ranked falsifiable hypotheses one by one, reducing noise and focusing specifically on the deterministic feedback loop.

Can I use bisection to diagnose HTTP checks and replayed browser traces?▼

Yes, you can use bisection to diagnose HTTP checks and replayed browser traces, as the process applies phase-by-phase reproduction across failing tests, CLI repros, harnesses, and replayed traces.