Memory Spring Cleaning

---
allowed-tools: Read, Write, Edit, Glob
argument-hint: [scope] | --claude-md | --documentation | --outdated-patterns | --implementation-sync
description: Clean and organize project memory files with implementation synchronization and pattern updates
---

# Memory Spring Cleaning

Clean and synchronize project memory with current implementation patterns: **$ARGUMENTS**

## Current Memory Context

- Memory files: !`find . -name "CLAUDE*.md" | wc -l` CLAUDE.md files in project
- Documentation: !`find . -name "README*" -o -name "*.md" | wc -l` total documentation files
- Last update: !`find . -name "CLAUDE.md" -exec stat -c "%y" {} \; 2>/dev/null | head -1 || echo "No CLAUDE.md found"`
- Implementation drift: Analysis of documented vs actual patterns

## Task

Execute comprehensive memory cleanup with implementation synchronization:

**Cleanup Scope**: Use $ARGUMENTS to focus on CLAUDE.md files, general documentation, outdated pattern identification, or implementation synchronization

**Memory Cleaning Framework**:
1. **Memory File Discovery** - Locate all CLAUDE.md and documentation files, assess hierarchy and organization, identify redundant content
2. **Implementation Analysis** - Compare documented patterns with actual code, identify implementation drift, assess accuracy gaps
3. **Pattern Validation** - Verify documented conventions, validate code examples, check dependency accuracy, assess technology stack alignment
4. **Content Optimization** - Remove outdated information, consolidate duplicate content, improve organization structure, enhance clarity
5. **Synchronization Updates** - Update development commands, refresh technology stack references, sync architectural patterns, validate workflows
6. **Quality Assurance** - Ensure consistency across files, validate markdown formatting, check link integrity, maintain version alignment

**Advanced Features**: Automated pattern detection, implementation drift analysis, cross-reference validation, documentation health scoring.

**Memory Health**: Content freshness metrics, accuracy validation, usage pattern analysis, maintenance scheduling recommendations.

**Output**: Cleaned and synchronized memory files with updated patterns, validated implementations, and maintenance recommendations.

Preview capped at 60,000 characters. Download the bundle for every original file and notice.