A Claude Code skill that analyzes macOS performance and provides actionable optimization recommendations.
Collects and analyzes:
- CPU and memory usage, swap pressure, compressor state
- Browser process bloat (Chrome, Brave, Safari, Firefox, Arc, Edge)
- Claude Code session count and resource usage
- Docker / OrbStack container runtimes
- Launch daemons and agents from third-party apps
- Disk space, system uptime, power settings
Then provides prioritized recommendations (HIGH / MEDIUM / LOW impact) with copy-pasteable commands.
npx skills add skalinets/mac_optimizer -g -yThe -g flag installs globally (available across all projects). The -y skips confirmation prompts.
git clone https://github.com/skalinets/mac_optimizer.git ~/work/mac_optimizer
ln -sfn ~/work/mac_optimizer ~/.claude/skills/mac-optimizergit clone https://github.com/skalinets/mac_optimizer.git
cp -r mac_optimizer ~/.claude/skills/mac-optimizerNote: After installation, restart Claude Code or start a new session for the skill to be available.
In any Claude Code session, type:
/mac-optimizer
Claude will run the analysis script and return a full performance report with recommendations.
You can also trigger it conversationally:
- "Check my Mac performance"
- "Why is my Mac slow?"
- "Optimize my system"
The analysis script can also be run directly in your terminal:
bash ~/.claude/skills/mac-optimizer/mac-optimize.shThis prints raw system metrics without the AI-powered analysis.
- macOS (tested on Apple Silicon only; Intel Macs have not been tested)
- Claude Code CLI installed
- No additional dependencies — uses only built-in macOS tools (
ps,vm_stat,sysctl,df,pmset, etc.)