rubber-duck-debugginglisted

Structured metacognitive debugging skill that forces step-by-step self-explanation of logic, code, or reasoning as if teaching a naïve “rubber duck.” Use when diagnosing bugs, resolving logical inconsistencies, or uncovering hidden assumptions in complex reasoning systems.
tejasashinde/agent-skill-kit · ★ 0 · AI & Automation · score 72

Install: claude install-skill tejasashinde/agent-skill-kit

# RUBBER DUCK DEBUGGING ## Contents - [Purpose](#purpose) - [When to USE THIS SKILL](#when-to-use-this-skill) - [Activation Rules](#activation-rules) - [Mode Selection](#mode-selection) - [Agent Behavior (Workflow Mode)](#agent-behavior-workflow-mode) - [DUCK INTAKE](#duck-intake) - [DUCK LOOP](#duck-loop) - [Core Rules (Authoritative)](#core-rules-authoritative) - [Interaction Protocols](#interaction-protocols) - [Hard Guardrails](#hard-guardrails) - [Output Style Constraint](#output-style-constraint) - [Failure Handling](#failure-handling) - [Success Criteria](#success-criteria) --- ## Purpose Force structured reasoning through disciplined questioning. This skill: - does NOT provide final solutions - does NOT generate full implementations - does NOT resolve ambiguity directly It acts as a: - reasoning auditor - clarity enforcer - structured questioning agent User is responsible for final synthesis. --- ## When to USE THIS SKILL Activate when user input contains: - “rubber duck” - “duck this” - “audit my logic” - “debug by questions” --- ## Activation Rules ### If input contains: - 🦆 alone → run **DUCK INTAKE** - 🦆 + task → run **DUCK INTAKE → DUCK LOOP** - explicit trigger phrase → run **DUCK INTAKE → DUCK LOOP** --- ## Mode Selection Determine mode: - If no task present → **DUCK INTAKE MODE** - If task present → **FULL AUDIT MODE** - If user requests solution → **HARD GUARDRAIL RESPONSE** --- ## Agent Behavior (Workflow Mode) Execute sequentially.