← ClaudeAtlas

safe-editlisted

Automatically backs up files, saves diffs, uses agents/skills, and ensures modular code (<200 lines) before any implementation. Use this skill for ALL code changes to ensure safe, reversible, and clean implementations.
aiskillstore/marketplace · ★ 329 · AI & Automation · score 82
Install: claude install-skill aiskillstore/marketplace
# Safe Edit Skill ## Overview Comprehensive workflow automation for safe, reversible, and clean code implementations. This skill ensures every code change follows best practices: automatic backups, diff tracking, agent utilization, and modular architecture enforcement. ## When to Use **ALWAYS activate this skill before ANY code implementation:** - Any file modification (create, update, delete) - Feature implementation - Bug fixes - Refactoring - UI/UX changes - API changes - Database schema updates **User triggers (automatic activation):** - Any implementation request without explicit backup mention - "implement", "add", "create", "fix", "update", "change" - Any code-related task ## Core Workflow ### Phase 1: Pre-Implementation Analysis ``` 1. Analyze task complexity and scope 2. Identify affected files and modules 3. Check if agents/skills can help 4. Plan modular architecture (if >200 lines) 5. Create TODO list for tracking ``` ### Phase 2: Backup & Safety ``` 1. Create timestamped backup in .backups/ 2. Generate git diff (save to /tmp/diffs/) 3. Document rollback commands 4. Verify backup integrity ``` ### Phase 3: Implementation ``` 1. Use appropriate agents/skills 2. Implement in modular chunks (<200 lines) 3. Follow existing patterns 4. Update TODO progress ``` ### Phase 4: Verification ``` 1. Generate final diff 2. Run type checking (if TypeScript) 3. Test build (if applicable) 4. Document changes ``` ## Backup Management ### Directory Structure ``` .backup