macos-cleaner_skill

Overview

This skill analyzes macOS disk usage and produces safe, actionable cleanup recommendations to reclaim space. It follows a safety-first philosophy: thorough analysis, clear impact descriptions, and explicit user confirmation before any deletions. The skill prioritizes conserving useful caches and avoiding destructive shortcuts.

How this skill works

The skill runs a staged workflow: quick diagnosis, deep analysis with a TTY-aware tool (Mole), category-by-category inspection, and conservative cleanup suggestions. It reports progress during long scans, explains consequences for each candidate, and outputs precise commands for the user to run manually after confirmation. Dangerous operations are flagged and require stepwise, per-item approval.

When to use it

• When a user reports low or shrinking disk space on macOS
• Before performing any automated cleanup on a Mac
• To identify large files, duplicate data, and leftover app remnants
• When assessing developer environment storage (Docker, Homebrew, npm, pip)
• When you need safe, explainable recommendations rather than one-click cleaning

Best practices

• Always run mo analyze in a TTY (use tmux) and be patient—home directory scans can take 5–10+ minutes
• Show what will happen if a file or folder is deleted for every recommendation
• Preserve caches that speed up downloads or builds if network or rebuild cost is high
• Never execute rm -rf or global prune commands without explicit, per-item user consent
• Provide cleanup commands for users to run themselves; do not auto-delete

Example use cases

• Diagnose a Mac with <10 GB free and identify largest reclaimable folders
• Find orphaned Application Support folders after apps were uninstalled
• List large files (>100 MB) and surface duplicates for user review
• Audit developer caches: Homebrew cache, npm/pip caches, Docker images and volumes with per-volume confirmation
• Generate step-by-step Mole tmux workflow and progress updates during long analyses