Cleanup

GabeDA Cleanup & Organization Expert

Purpose

This skill analyzes and organizes project folders to maintain clean, well-structured codebases. It understands project conventions, identifies redundancy, consolidates accumulator files, and removes obsolete content.

Key Capabilities:

• Analyze current folder/project state
• Detect redundant and duplicate content across files ⭐ NEW
• Suggest document consolidation and merging ⭐ NEW
• Move files to proper locations per project conventions
• Append content to accumulator/living documents
• Identify and remove obsolete files
• Generate detailed operation and merge suggestion reports

When to Invoke

Use this skill when:

• Project has accumulated duplicate/scattered files
• Multiple files contain redundant or overlapping content ⭐ NEW
• Need suggestions for document consolidation/merging ⭐ NEW
• Files need reorganization per documentation standards
• Content should be consolidated into living documents
• Obsolete files need identification and removal
• Need audit trail of file operations

Input required: Path(s) to folders needing cleanup

Workflow

1. Analyze Current State

Understand project structure:

• Read project documentation standards (CLAUDE.md, DOCUMENTATION_STANDARD.md)
• Identify living documents (10 accumulators in ai/ and docs/)
• Map current file locations and relationships
• Detect redundancy patterns

Generate state report:

• Total files by type/category
• Misplaced files (wrong folder)
• Duplicate/redundant content
• Accumulator opportunities (content that should be appended)
• Obsolete file candidates

2. Detect Redundancy (⭐ NEW)

Analyze files for redundant content:

Content Similarity Analysis:

• Compare file contents pairwise
• Calculate similarity scores (0-100%)
• Identify duplicate/near-duplicate clusters
• Group files by content similarity

Pattern Recognition:

• Detect versioned files (file_v1.md, file_v2.md, file_final.md)
• Detect dated files (notes_2025_10.md, notes_2025_11.md)
• Detect topic variants (auth.md, authentication.md, login_guide.md)

Topic Modeling:

• Extract topics from file headings
• Group files covering same topics
• Identify fragmented information

Reference Analysis:

• Count references to each file
• Identify authoritative sources (heavily referenced)
• Find orphan files (unreferenced duplicates)

Redundancy Scoring:

• Assign redundancy score (0-10) to file groups
• HIGH (9-10): Immediate consolidation recommended
• MEDIUM (6-8): User review suggested
• LOW (3-5): Optional optimization

Generate merge suggestions:

• Present redundancy groups by priority
• Propose consolidation strategies (MERGE/CONSOLIDATE/HIERARCHICAL/APPEND)
• Estimate impact (file reduction, clarity improvement)
• Request user approval before proceeding

Reference: cleanup/references/redundancy_detection.md

Template: cleanup/assets/templates/merge_suggestion_template.md

3. Categorize Files

Decision tree for each file:

FILE → Is it a living document?
       ├─ YES → Check if properly named/located
       │        ├─ YES → Skip (already correct)
       │        └─ NO → Flag for rename/move
       │
       └─ NO → Should content go into living doc?
               ├─ YES → Flag for append + delete
               │
               └─ NO → Is file obsolete?
                       ├─ YES → Flag for deletion
                       └─ NO → Is file in wrong folder?
                               ├─ YES → Flag for move
                               └─ NO → Skip (correct)

Reference: cleanup/references/file_categorization_rules.md

4. Plan Operations

Group operations by type:

1. MERGE - Consolidate redundant files (⭐ NEW)
2. MOVE - Files to relocate (with destination)
3. APPEND - Content to merge into living docs
4. DELETE - Obsolete files to remove
5. RENAME - Files needing name standardization

Safety checks:

• Verify destinations exist
• Check for overwrite conflicts
• Validate accumulator file format compatibility
• Confirm deletion safety (no external references)
• Ensure unique content preserved during merges (⭐ NEW)

5. Execute with User Approval

Present operation plan:

• Show all proposed operations in clear table format
• Highlight potential risks (overwrites, deletions)
• Explain rationale for each operation

Request approval:

• CRITICAL: NEVER execute operations without explicit user approval
• User reviews plan and confirms: "proceed", "skip deletions", "cancel", etc.

Execute approved operations:

• Perform operations in safe order: MERGE → MOVE → APPEND → DELETE
• Log each operation with timestamp
• Handle errors gracefully (report, don't abort entire batch)

6. Generate Report

Create operation report using template: cleanup/assets/templates/cleanup_report_template.md

Report sections:

• Summary: Total operations by type, files affected
• Detailed log: Each operation with source/destination/timestamp
• Errors: Any failures encountered
• State comparison: Before/after folder structure
• Recommendations: Further cleanup opportunities

Save report: cleanup_report_YYYYMMDD_HHMM.md in project root or user-specified location

Living Documents (10 Accumulators)

These files accumulate content - APPEND, don't replace:

ai/ Folder (AI Skill Accumulators)

1. ai/CHANGELOG.md - All code changes
2. ai/ISSUES.md - Bug fixes and resolutions
3. ai/PROJECT_STATUS.md - Sprint status and metrics
4. ai/FEATURE_IMPLEMENTATIONS.md - New features added
5. ai/TESTING_RESULTS.md - Test execution results
6. ai/SKILLS_MANAGEMENT.md - AI skills evolution
7. ai/testing/TEST_MANIFEST.md - Complete test catalog

docs/ Folder (Human-Readable)

8. ai/architect/ARCHITECTURE_DECISIONS.md - Design decisions (ADRs)
9. ai/guides/NOTEBOOK_IMPROVEMENTS.md - Notebook refactoring log
10. ai/planning/FUTURE_ENHANCEMENTS.md - Future TODOs

Reference: cleanup/references/accumulator_patterns.md

File Organization Rules

Standard locations per DOCUMENTATION_STANDARD.md:

Reference: cleanup/references/folder_structure_guide.md

Obsolescence Detection

Files are likely obsolete if:

• Contains "OLD", "BACKUP", "DEPRECATED" in name
• Duplicates content in living documents
• Replaced by newer versioned file
• References outdated architecture (pre-refactoring)
• Not referenced by any other file in project
• Creation date > 90 days + no recent modifications

Manual review required before deletion:

• Configuration files
• Data files (CSVs, JSONs)
• Files > 100 lines
• Any file with "IMPORTANT" or "CRITICAL" markers

Reference: cleanup/references/obsolescence_criteria.md

Safety Guidelines

CRITICAL RULES:

1. ALWAYS get user approval before executing operations
2. NEVER delete without showing what will be deleted
3. NEVER overwrite without warning about conflicts
4. ALWAYS preserve original content (copy before transform)
5. ALWAYS generate operation report for audit trail

Dry-run mode: Present plan WITHOUT executing (default behavior)

Execution mode: Only after explicit user approval

Example Usage

User: "Clean up the ai/ folder"

Cleanup skill response:

## Analysis Complete: ai/ Folder

**Current State:**
- 156 files total
- 12 living documents (correct)
- 8 misplaced files detected
- 3 files should append to accumulators
- 5 obsolete files identified

**Proposed Operations:**

### MOVE (8 files)
1. ai/old_feature_spec.md → ai/specs/archive/
2. ai/test_results_20251020.md → ai/testing/archive/
   [... 6 more]

### APPEND (3 files)
1. ai/bug_fix_notes.md → ai/ISSUES.md (append content, then delete source)
2. ai/new_model_added.md → ai/FEATURE_IMPLEMENTATIONS.md
3. ai/sprint4_update.md → ai/PROJECT_STATUS.md

### DELETE (5 files)
1. ai/BACKUP_old_spec.md (obsolete - content in specs/)
2. ai/temp_notes.txt (temporary file, 92 days old)
   [... 3 more]

**Approve operations? (respond: 'proceed', 'skip deletions', 'cancel')**

More examples: cleanup/assets/examples/cleanup_scenarios.md

Example Usage with Redundancy Detection (⭐ NEW)

User: "Clean up the docs/ folder and suggest any document consolidation"

Cleanup skill response:

## Analysis Complete: docs/ Folder

**Current State:**
- 25 documentation files analyzed
- 3 living documents (correct)
- 1 README (correct)

**Redundancy Detection Results:**

### HIGH Redundancy (Score: 9/10)
**Group 1: Setup Guides (3 files)**
- setup_guide_v1.md (150 lines, modified 2025-09-15)
- setup_guide_v2.md (180 lines, modified 2025-10-01)
- setup_guide_final.md (200 lines, modified 2025-10-25) ← Most recent

Content similarity: 85% overlap
Suggestion: MERGE → Keep setup_guide_final.md, delete v1 and v2

### MEDIUM Redundancy (Score: 7/10)
**Group 2: Authentication Documentation (2 files)**
- api_authentication.md (120 lines, referenced by 3 files)
- auth_endpoints.md (95 lines, not referenced)

Content similarity: 70% overlap
Topic: Both cover authentication/login endpoints
Suggestion: CONSOLIDATE → Merge into single authentication.md

### Proposed Consolidation Operations:

**MERGE (Group 1):**
1. Keep: setup_guide_final.md
2. Extract unique content from v1 and v2 (if any)
3. Delete: setup_guide_v1.md, setup_guide_v2.md
4. Impact: 3 files → 1 file

**CONSOLIDATE (Group 2):**
1. Create: authentication.md
2. Merge content from both files
3. Update 3 references to point to new file
4. Delete source files
5. Impact: 2 files → 1 file

**Total Impact:**
- File reduction: 25 → 20 files (5 files consolidated, 20% reduction)
- Improved clarity: Single authoritative source for each topic
- Easier maintenance: Fewer files to keep updated

**Do you approve these consolidations?**
(respond: 'approve-all', 'approve group 1 only', 'modify', 'skip')

User response: "approve-all"

Cleanup skill executes consolidation + generates report with before/after comparison

Quality Standards

Cleanup operations must meet:

✅ Correctness (25%) - Files moved to proper locations per standards ✅ Safety (25%) - No data loss, user approval obtained ✅ Completeness (20%) - All redundancy/obsolescence addressed ✅ Auditability (20%) - Detailed operation report generated ✅ Reversibility (10%) - Operations can be undone if needed

Minimum score: 8.0/10 before executing operations.

Related Skills

Use other skills for:

• executive - Decide project-level organization strategy
• architect - Understand code architecture for proper categorization
• business - Identify which documents are accumulator vs reference
• insights - Organize notebook files and analysis outputs
• marketing - Organize marketing assets and templates
• ux-design - Organize design assets and wireframes

Bundled Resources

References:

• file_categorization_rules.md - Decision logic for categorizing files
• accumulator_patterns.md - How to append to living documents
• folder_structure_guide.md - Standard project organization
• obsolescence_criteria.md - Rules for identifying obsolete files
• redundancy_detection.md - ⭐ NEW: Detect and consolidate redundant content

Templates:

• cleanup_report_template.md - Standard operation report format
• operation_plan_template.md - Pre-execution approval format
• merge_suggestion_template.md - ⭐ NEW: Document merge suggestions

Examples:

• cleanup_scenarios.md - Real cleanup examples with rationale

Version: 1.1.0 Last Updated: 2025-10-30 Skill Type: Project Organization & Redundancy Detection