hunt

Diagnose root causes of errors, crashes, and regressions using repro steps and source traces.

8|Updated Jan 19, 2026
One-click install
npx skills add https://github.com/OiAnthony/.agents --skill hunt-oianthony
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: hunt
Source: https://github.com/OiAnthony/.agents/tree/main/skills/hunt
Command: npx skills add https://github.com/OiAnthony/.agents --skill hunt-oianthony

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill helps diagnose errors, crashes, regressions, and broken behavior by finding the root cause before any fix is applied. It is designed for situations where something used to work, now fails, or behaves differently than expected.

Core Features & Use Cases

  • Root-cause analysis first: Requires a specific, testable hypothesis before touching code.
  • Evidence-driven debugging: Emphasizes repro steps, logs, source trace, runtime state, and targeted verification.
  • Regression and failure recovery: Handles failing tests, screenshot-reported defects, stale cache issues, and timing-sensitive bugs.
  • Scope control: Prevents symptom-level patches, encourages blast-radius checks, and surfaces when to hand off after repeated failed hypotheses.

Quick Start

Use the hunt skill to reproduce the failure, identify the root cause in one sentence, and verify the fix with a targeted check before making changes.

Frequently Asked Questions about hunt

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

FAQPage Schema
How do I find the root cause of failing tests and regressions before applying a fix?

Root cause analysis for failing tests and regressions requires gathering a reproducible check, executing a source trace, collecting runtime evidence, and running targeted verification before applying any code fix.

What is the best way to debug screenshot-reported defects and runtime failures?

Debugging screenshot-reported defects and runtime failures relies on evidence-driven debugging, which emphasizes reproducing the failure, tracing the source, and checking the blast radius to prevent symptom-level patches.

How do I stop applying symptom-level patches when diagnosing broken behavior?

To stop applying symptom-level patches when diagnosing broken behavior, enforce scope control by requiring a specific, testable hypothesis and targeted verification before touching any code.

When should I hand off debugging after repeated failed hypotheses for a regression?

You should hand off debugging after repeated failed hypotheses when the root cause analysis surfaces repeated dead ends, indicating the regression or runtime failure requires a different investigative scope.

Can I use evidence-driven debugging for timing-sensitive bugs and stale cache issues?

Yes, evidence-driven debugging applies to timing-sensitive bugs and stale cache issues by utilizing runtime state, repro steps, and logs to verify the exact conditions causing the regression.