skills.rest
yarma-tech
yarma-techCOMMUNITY
Explore all skills by this author →
GitHub

diagnose

Guide a disciplined loop from reproduction to regression testing for hard bugs.

Updated Aug 27, 2026
Search Tags:#debugging#instrumentation#diagnose#reproduction#regression-testing#hypothesising
One-click install
npx skills add https://github.com/yarma-tech/agent-admin-qonto --skill diagnose-yarma-tech
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: diagnose
Source: https://github.com/yarma-tech/agent-admin-qonto/tree/main/.agents/skills/diagnose
Command: npx skills add https://github.com/yarma-tech/agent-admin-qonto --skill diagnose-yarma-tech

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Disciplined diagnosis loop for hard bugs and performance regressions. Reproduce → minimise → hypothesise → instrument → fix → regression-test. Use when user says "diagnose this" / "debug this", reports a bug, says something is broken/throwing/failing, or describes a performance regression.

Core Features & Use Cases

  • Structured feedback loop: a phased workflow that guides reproducibility, hypothesis testing, instrumentation, and verification.
  • Phase-driven diagnosis: build a repeatable loop from reproduction to regression testing, with guardrails and clear exit criteria.
  • Practical guidance: provides concrete steps for creating deterministic signals (tests, curl against a server, CLI invocations, log-based tracing, or replaying traces) and for collecting evidence.

Quick Start

Initiate a disciplined bug-diagnosis session and guide the agent through reproducing, minimising, hypothesising, instrumenting, fixing, and regression-testing the issue.

Frequently Asked Questions about diagnose

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

FAQPage Schema
How do I debug hard bugs and performance regressions systematically?▼

Debug hard bugs systematically using a disciplined loop that reproduces, minimises, hypothesises, instruments, fixes, and regression-tests the issue. This structured workflow enforces phased feedback loops and evidence collection to ensure verifiable fixes.

What is the best way to reproduce a bug and collect diagnostic evidence?▼

Reproduce a bug and collect evidence by creating deterministic signals such as tests, CLI invocations, curl commands against servers, log-based tracing, or replaying traces. This approach builds a repeatable loop from reproduction to regression testing.

How do you form and test a hypothesis when diagnosing a broken behavior?▼

Diagnose broken behavior by applying a phase-driven workflow that requires explicit hypotheses and structured evidence collection. You instrument the code to generate measurable signals that verify or falsify each hypothesis before applying a fix.

Do I need a mental model of the codebase to diagnose performance regressions?▼

Yes, diagnosing performance regressions requires a clear mental model of the codebase, instrumentation, and measurable signals. This prerequisite ensures you can effectively navigate the phased feedback loop and interpret the collected diagnostic evidence.

When should I use a structured debugging loop instead of ad hoc debugging?▼

Use a structured debugging loop when users report hard bugs, broken behavior, throwing errors, or performance regressions that resist ad hoc debugging. The enforced workflow with guardrails and clear exit criteria prevents premature fixes by verifying hypotheses.

Related Skills

Explore More Software Engineering Skills→
BRANDNEWSHVTBRANDNEWSHVT

codebase-mapping

Map directory structure and trace data flows across codebase modules.

Official
Advanced
SjdjdiejdrirhdkjejSjdjdiejdrirhdkjej

regression-test-generator

Generate regression tests from bug reports and detected test frameworks.

Community
Advanced
jwilgerjwilger

event-modeling

Guide structured event modeling workshops for event-sourced systems.

Community
Advanced
yash-gargyash-garg

diagnosing-bugs

Diagnose software bugs and performance regressions through reproducible feedback loops.

Community
Advanced