refactor-code

name	refactor-code
description	Safely improve code quality through test-driven refactoring with automatic validation and rollback. Use during quality review to reduce technical debt while preserving behavior.
acceptance	[{"config_validated":"Configuration allows refactoring and prerequisites verified"},{"opportunities_identified":"Refactoring opportunities identified and prioritized by risk"},{"refactorings_applied":"Selected refactorings applied incrementally with test validation"},{"log_created":"Refactoring log documenting all changes with rationale"},{"tests_passing":"All tests passing after refactoring with no behavioral changes"}]
inputs	{"task_file":{"type":"string","required":true,"description":"Path to task specification with quality findings"},"aggressiveness":{"type":"enum","required":false,"description":"Refactoring scope: conservative, moderate, aggressive","default":"moderate"},"allow_refactoring":{"type":"boolean","required":false,"description":"Override config to allow/disallow refactoring"}}
outputs	{"refactorings_applied":{"type":"number","description":"Number of refactorings successfully applied"},"refactorings_failed":{"type":"number","description":"Number of refactorings that failed and were rolled back"},"files_modified":{"type":"array","description":"List of files modified during refactoring"},"quality_impact":{"type":"object","description":"Before/after quality metrics showing improvements"}}
telemetry	{"emit":"skill.refactor-code.completed","track":["refactorings_applied","refactorings_failed","refactorings_skipped","success_rate","files_modified_count","quality_improvement_percent","duration_ms"]}

Refactor Code Skill

Purpose

Safely improve code quality through automated refactoring while maintaining behavioral correctness. This skill identifies refactoring opportunities from quality findings and applies them incrementally with continuous test validation.

Core Capabilities:

• Test-driven refactoring (run tests after each change)
• Automatic rollback on test failures
• Risk-based prioritization (P0-P3, Low/Medium/High risk)
• Incremental application (one refactoring at a time)
• Full traceability (log all changes with rationale)
• Quality metrics tracking (before/after comparison)

Key Innovation: Safety-first approach ensures refactorings never break functionality. Each change is validated with tests before proceeding.

Prerequisites

• Task status is "Review"
• Quality assessment has been run (quality gate exists)
• Tests exist and are currently passing
• Configuration allows refactoring (quality.allowRefactoring: true)

---

Workflow

Step 0: Load Configuration and Validate Prerequisites

Action: Verify refactoring is allowed and safe to proceed.

Load configuration: Read .claude/config.yaml for refactoring settings

See: references/templates.md#configuration-format for complete config structure

Check prerequisites:

1. Verify allowRefactoring: true in config (or input override)
2. Verify task status is "Review"
3. Verify quality assessment exists (.claude/quality/gates/{task-id}-gate.yaml)
4. Verify tests exist and are passing

Get refactoring scope: Ask user to choose Conservative/Moderate/Aggressive or Skip

See: references/templates.md#user-decision-refactoring-scope for prompt format

Halt if:

• Configuration disables refactoring and no input override
• No tests exist (can't validate safety)
• Tests are currently failing
• Quality assessment not yet run
• User chooses to skip

Output: Prerequisites validation summary (config, tests, scope)

See: references/templates.md#step-0-output-template for format