What problem does it solve?
This Skill helps when an Apple Silicon macOS dev machine gets into a failure pattern where low disk space and memory pressure escalate into freezes, reboots, and watchdog-timeout panics. It prevents “clean-up” work from becoming guesswork by combining signal triage, safety-guarded cleanup planning, and an explicit user approval step.
Core Features & Use Cases
- Signal triage playbook: classifies incidents (disk pressure, memory pressure, kernel panic, JetsamEvent vm-compressor-space-shortage) and selects an appropriate cleanup tier.
- Interactive cleanup UI (local-only): renders scan results as a categorized HTML checklist with tooltips and a single selection JSON as the source of truth.
- Sanctioned apply step with validation: applies only the user-selected commands and blocks unsafe/protected paths using Mole-style safety checks.
- Proactive LaunchAgent alerter: sends CRITICAL-only notifications with hysteresis, cooldown, and a 7-day calibration window to avoid alert fatigue.
Quick Start
Tell your AI agent: “I’m getting ‘Mac is full’ and my disk is under 20%—run the maintaining-macos-health triage, open the cleanup plan UI, and apply only what I select after I confirm go.”