scbe-disk-management

name	scbe-disk-management
description	Audit disk usage, identify space hogs, and safely clean up caches/temp files across the SCBE-AETHERMOORE repo and user profile. Use when asked about disk space, storage cleanup, finding large files, or freeing space.

SCBE Disk Management

Wraps the three existing PowerShell scripts in scripts/system/ for disk auditing and safe cleanup.

Scripts

Script	Purpose	Output
system_disk_audit.ps1	Audit SCBE repo directory sizes + C: drive free/used	artifacts/system-audit/disk_audit.json + .md
system_profile_disk_audit.ps1	Audit entire user profile (C:\Users\issda) top-level dirs	artifacts/system-audit/profile_disk_audit.json + .md
system_cleanup_safe.ps1	Remove safe-to-delete caches (pytest, hypothesis, pycache)	Console output

Operations

1. Repo Disk Audit

Scans targeted SCBE directories (src, docs, tests, node_modules, dist, artifacts, training, training-data, .hypothesis, .pytest_cache) plus all top-level children ranked by size.

powershell -ExecutionPolicy Bypass -File scripts/system/system_disk_audit.ps1

Parameters:

• -RepoPath — repo root (default: C:/Users/issda/SCBE-AETHERMOORE)
• -TopN — number of largest children to show (default: 25)
• -OutDir — output subdirectory (default: artifacts/system-audit)

Reports C: drive used/free GB, targeted path sizes, and largest top-level folders. Writes both JSON and Markdown.

2. Profile Disk Audit

Scans all top-level directories under the user profile to find the biggest space consumers system-wide.

powershell -ExecutionPolicy Bypass -File scripts/system/system_profile_disk_audit.ps1

Parameters:

• -Root — scan root (default: C:/Users/issda)
• -TopN — number of results (default: 30)
• -OutDir — output subdirectory (default: artifacts/system-audit)

3. Safe Cleanup (dry-run first)

Removes safe-to-delete cache directories. Always run dry-run first.

# Dry-run (shows what WOULD be deleted, deletes nothing)
powershell -ExecutionPolicy Bypass -File scripts/system/system_cleanup_safe.ps1

# Actually delete
powershell -ExecutionPolicy Bypass -File scripts/system/system_cleanup_safe.ps1 -Apply

Cleanup targets (only within the repo):

• .pytest_cache
• .hypothesis/examples
• artifacts/pytest_tmp
• __pycache__ (root, src, tests)

Parameters:

• -RepoPath — repo root (default: C:/Users/issda/SCBE-AETHERMOORE)
• -Apply — switch to actually delete (omit for dry-run)

Recommended Workflow

1. Audit first: Run repo audit to see current sizes
2. Profile audit: If C: drive is low, run profile audit to find what's eating space outside the repo
3. Dry-run cleanup: Run cleanup without -Apply to review targets
4. Apply cleanup: If safe, re-run with -Apply
5. Re-audit: Confirm space was freed

Reading Reports

After auditing, reports are at:

• artifacts/system-audit/disk_audit.json — repo audit (machine-readable)
• artifacts/system-audit/disk_audit.md — repo audit (human-readable)
• artifacts/system-audit/profile_disk_audit.json — profile audit (machine-readable)
• artifacts/system-audit/profile_disk_audit.md — profile audit (human-readable)

Read the .md files for a quick summary, or parse the .json for automation.

Safety Rules

• Never delete node_modules without confirmation — requires npm install to restore
• Never delete training-data/ or training/ — irreplaceable SFT pairs
• Never delete .git/ — repository history
• Never delete dist/ without rebuilding — breaks npm package
• The cleanup script only targets expendable caches that regenerate automatically
• Always dry-run before applying