cleanup

User Input

$ARGUMENTS

Options: dry-run (preview only), force (execute deletions)

Task

Execute data retention cleanup according to the privacy policy.

Steps

1. Parse Arguments:

   • dry-run or empty: Preview deletions without executing
   • force: Execute actual deletions

2. Cleanup Paid Quiz Responses (24h after PDF delivery):

   bashCopy

   cd backend
   python scripts/cleanup_paid_quiz.py --dry-run
   
   # Expected: Delete quiz_responses where pdf_delivered_at < NOW() - 24h
   

3. Cleanup Unpaid Quiz Responses (7 days after creation):

   bashCopy

   python scripts/cleanup_unpaid_quiz.py --dry-run
   
   # Expected: Delete quiz_responses where created_at < NOW() - 7d AND payment_id IS NULL
   

4. Cleanup Meal Plans (90 days after creation):

   bashCopy

   python scripts/cleanup_meal_plans.py --dry-run
   
   # Expected: Delete meal_plans where created_at < NOW() - 90d
   

5. Cleanup PDFs from Vercel Blob (91 days = 90d + 24h grace):

   bashCopy

   python scripts/cleanup_pdfs.py --dry-run
   
   # Expected: Delete blobs where created_at < NOW() - 91d
   

6. Cleanup Expired Magic Links (24h after creation):

   bashCopy

   python scripts/cleanup_magic_links.py --dry-run
   
   # Expected: Delete magic_link_tokens where created_at < NOW() - 24h
   

7. Cleanup Email Blacklist (90-day TTL):

   bashCopy

   python scripts/cleanup_blacklist.py --dry-run
   
   # Expected: Delete email_blacklist where created_at < NOW() - 90d
   

8. Generate Deletion Report:

   Copy

   ✅ Data Retention Cleanup Report
   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
   Mode: DRY RUN (no deletions executed)
   
   Quiz Responses (Paid):
   📊 Found: 145 records eligible for deletion
   📅 Oldest: 25 days since PDF delivery
   🗑️  Would delete: 145 records
   
   Quiz Responses (Unpaid):
   📊 Found: 23 records eligible for deletion
   📅 Oldest: 15 days old
   🗑️  Would delete: 23 records
   
   Meal Plans:
   📊 Found: 8 records eligible for deletion
   📅 Oldest: 95 days old
   🗑️  Would delete: 8 records
   
   PDFs (Vercel Blob):
   📊 Found: 8 blobs eligible for deletion
   💾 Space to reclaim: 4.2 MB
   🗑️  Would delete: 8 blobs
   
   Magic Links:
   📊 Found: 67 expired tokens
   🗑️  Would delete: 67 records
   
   Email Blacklist:
   📊 Found: 3 expired entries
   🗑️  Would delete: 3 records
   
   Total Deletions: 254 records + 8 blobs
   Space Reclaimed: 4.2 MB
   
   ⚠️ This was a DRY RUN. Use '/cleanup force' to execute.
   

9. Execute Deletions (if --force):

   • Run all cleanup scripts with --force flag
   • Log all deletions to Sentry for audit trail
   • Generate post-cleanup report

10. Audit Logging:

    bashCopy

    # Log to Sentry
    python -c "
    import sentry_sdk
    sentry_sdk.capture_message(
        'Data cleanup executed',
        level='info',
        extra={
            'quiz_deleted': 168,
            'meal_plans_deleted': 8,
            'pdfs_deleted': 8,
            'magic_links_deleted': 67,
            'blacklist_deleted': 3
        }
    )
    "
    

Example Usage

/cleanup              # Dry run - preview deletions
/cleanup dry-run      # Same as above
/cleanup force        # Execute actual deletions

Exit Criteria

• All cleanup scripts executed
• Deletion counts reported
• Audit logs created (for force mode)
• Storage space reclaimed calculated

Safety Notes

• Always run dry-run first to verify deletions
• Deletions are permanent - no undo
• Audit trail required - all deletions logged to Sentry
• Compliance - retention policy must be enforced