cloud-storage-local-storage-management

name	cloud-storage-local-storage-management
description	Audit, sort, route, and back up large Windows workspaces across local disks and cloud-sync folders without recursive mirror failures. Use when the user wants the filesystem cleaned up, cloud storage organized, backup targets chosen by free space and route quality, or hardware/storage bottlenecks identified before moving data.

Cloud Storage and Local Storage Management

Use this skill when the user says some version of:

• "push everything to cloud storage"
• "my filesystem is messed up"
• "sort my local files"
• "find the best backup route"
• "clean up storage"
• "optimize hardware/storage"

This is a storage-management skill, not a fake security ritual.

• Do not hardcode birthdays, names, or simple codes as "admin security".
• Real admin actions should happen through the normal command-escalation flow.
• Default to scans, plans, and dry-run backup routes before destructive cleanup.

What this skill does

1. Scan the machine layout:
   • local drives
   • cloud roots
   • biggest folders
   • biggest files
   • known cache/problem zones
2. Rank backup routes by:
   • free space
   • whether the cloud root already exists
   • whether the route will recurse into itself
3. Emit a sort plan instead of guessing moves.
4. Run a profile backup with sane excludes when the route is safe.

Use the bundled scripts in this order

1. Scan

pwsh -File skills/cloud-storage-local-storage-management/scripts/storage_route_scan.ps1

Outputs JSON + Markdown into artifacts/storage-management/.

2. Build a sort plan

pwsh -File skills/cloud-storage-local-storage-management/scripts/storage_sort_plan.ps1

This converts the scan into a deterministic cleanup and backup plan.

3. Run a backup lane

Dry run first:

pwsh -File skills/cloud-storage-local-storage-management/scripts/backup_profile.ps1

Apply only after the route is acceptable:

pwsh -File skills/cloud-storage-local-storage-management/scripts/backup_profile.ps1 -Apply

Operating rules

• Prefer the smallest safe change that improves structure.
• Never back up a profile into a cloud root without excluding that root itself.
• Treat cloud folders inside the source tree as recursion hazards.
• Treat app caches, temp stores, package caches, and mail-bridge stores as separate from user documents.
• If the target cloud drive is nearly full, stop and reroute instead of forcing the copy.

Hardware and route logic

The route decision is based on:

• target exists
• target drive free GB
• target path stability
• whether the route is likely to self-nest
• whether the target is a content folder or a sync-engine cache

Good routes:

• Dropbox\machine-backup\...
• Drive\machine-backup\...
• OneDrive\machine-backup\...
• another local disk outside the source tree

Bad routes:

• copying into a folder that is already inside the source without excluding it
• backing up large app caches into a nearly full sync target
• using Creative Cloud or mail stores as generic bulk backup targets

Admin behavior

If a scan or backup needs higher privileges:

• use the normal escalated command flow
• explain why
• keep the command narrow

Do not simulate "security" with a shared secret embedded in the skill.

Resources

scripts/

• storage_route_scan.ps1: inspect the current storage layout and emit a ranked route report
• storage_sort_plan.ps1: turn the latest scan into a cleanup/backup plan
• backup_profile.ps1: dry-run or execute a safe profile backup with recursion and capacity guards