debug

Pre-computed context

Current branch: !git branch --show-current 2>/dev/null || echo "unknown" Recent commits: !git log --oneline -10 2>/dev/null || echo "no commits" Working tree status: !git status --porcelain 2>/dev/null | head -10 || echo "clean"

Variables

Arguments: $ARGUMENTS

Purpose

Hard bugs are won or lost in Phase 1. Without a fast, deterministic, agent-runnable signal that says "bug present / bug fixed", every later phase is guessing — and most failed debugging sessions fail because the engineer skipped straight to hypothesising without building a loop.

This skill enforces the discipline. Six phases, each with a clear gate before the next. The middle three (hypothesise → instrument → fix) are mechanical once Phase 1 is solid; the bookends (loop, cleanup) are the load-bearing work.

Scope boundary — this skill starts from an observed failure: UI behaving wrong, a log line that should not appear, a performance regression, a screenshot of a bug, a production symptom. Its first job is to construct a reproduction loop. If the symptom is already a failing test with no reproduction gap, you do not need this skill — cycle that test directly (reproduce → fix → retest → regression). What /debugging:debug adds over a bare fix loop is a critical edge case: if no correct test seam exists, that absence IS the finding — filed as an architectural recommendation, not a forced test in the wrong place.

Adapting to your environment (graceful degrade)