Diagnose software errors by testing hypotheses with targeted instrumentation and evidence.
npx skills add https://github.com/einargudnig/dotfiles --skill hunt-einargudnig
Please help me install this Agent Skill. Skill: hunt Source: https://github.com/einargudnig/dotfiles/tree/main/claude/skills/hunt Command: npx skills add https://github.com/einargudnig/dotfiles --skill hunt-einargudnig
This Skill enforces a disciplined diagnostic workflow to find the true root cause of errors, crashes, unexpected behavior, and failing tests before any code changes are made, preventing symptom-only patches that introduce new bugs.
State a one-sentence, testable hypothesis identifying the exact file:line and add one targeted instrument to prove or disprove it before proposing any fix.
High-intent search queries and answers about installing and using this skill.
To find the root cause of failing tests, state a one-sentence testable hypothesis naming the exact file:line, add minimal targeted instrumentation to verify it, and gather two independent evidence points before changing code.
Hypothesis-first debugging requires naming a specific file, function, or line in a one-sentence testable hypothesis before attempting fixes, preventing symptom-only patches by forcing targeted evidence collection and verification.
Diagnose CI failures and environment mismatches by explicitly enumerating environment differences, using grep to confirm suspected identifiers, and adding targeted assertions or logs to validate the root cause hypothesis.
The best way to debug crashes and rendering issues is to apply disciplined diagnosis: add the smallest possible instrument to prove or disprove a hypothesis, confirm with independent evidence, and only then implement and test the fix.
Stop debugging and escalate after repeated failed hypotheses by enumerating environment differences, confirming identifiers with grep, and reporting a concise outcome that names the root cause, fix, and confirmation evidence.
Yes, you can reproduce a failing CI test by naming the exact suspected file:line, adding a minimal assertion or log as targeted instrumentation, and confirming with two independent evidence points before testing a fix.
mvillmowParse test logs to extract and categorize failures by type.
allexcdDiagnose, reproduce, and fix code defects from bug reports or error logs.
treejhDiagnose root causes of errors and failing tests with evidence-based hypotheses.
akvoGuide structured root cause investigation through hypothesis testing and evidence gathering.
