syscheck

One command. Full system pulse. Zero dependencies.

Why

You open your laptop. The fans spin up. Something is eating CPU. Something else grabbed a port you need. An LLM is chewing through your RAM. You don't want to run five different commands to figure it out.

syscheck gives you the full picture in one shot.

What it shows

  ┌─┐┬ ┬┌─┐┌─┐┬ ┬┌─┐┌─┐┬┌─   Apple M5 Max
  └─┐└┬┘└─┐│  ├─┤├┤ │  ├┴┐   May 01 12:17 · up 7 days, 3:19
  └─┘ ┴ └─┘└─┘┴ ┴└─┘└─┘┴ ┴
  ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
  🧠 CPU  ███████░░░░░░░░░░░░░  35%  usr:26.3% sys:8.5%
  💾 RAM  ████████░░░░░░░░░░░░  42%  15.4G/36.0G
  🎮 GPU  ███░░░░░░░░░░░░░░░░░  16%  render:16% vram:1.2G
  💿 DSK  ░░░░░░░░░░░░░░░░░░░░   1%  12Gi/1.8Ti
  🔄 SWP  ████████░░░░░░░░░░░░  44%  1363M/3072M
  ⚖️  LD   2.52 / 18 cores
  🌍 NET  ↓79.2G ↑15.6G  🔋 BAT  100% ⚡

  🤖 LLM  total cpu:10.7%  ram:1.6G
  ──────────────────────────────────────────────
  │ Claude       pid:62939  cpu:  0.0%  ram:946M
  │ Claude       pid:64850  cpu: 10.7%  ram:729M

  🔥 HOT
  ──────────────────────────────────────────────
  │ python             pid:62950  cpu:113.3%  mem:2.7%
  │ WindowServer       pid:403    cpu: 25.1%  mem:0.5%

  📊 MEM  (use --all for more)
  ──────────────────────────────────────────────
  │ python             pid:62950    1009M  cpu:105.4%
  │ WallpaperAerial    pid:734       988M  cpu:0.0%
  │ claude             pid:62939     946M  cpu:0.1%

  🌐 NET  (system ports hidden)
  ──────────────────────────────────────────────
  │ :8325   python3.1        pid:40256
  │ :52564  python3.1        pid:62950

  ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

All output is color coded with progress bars. Green under 50%, yellow 50 to 80%, red above 80%.

Section	Details
🧠 CPU	Total CPU usage with user/system split
💾 RAM	Memory usage with used/total breakdown
🎮 GPU	Apple Silicon GPU utilization, renderer %, VRAM in use
💿 DSK	Root disk usage
🔄 SWP	Swap usage (only shown when swap is active)
⚖️ LD	Load average vs core count
🌍 NET	Network bytes in/out since boot
🔋 BAT	Battery percentage and charging state
🤖 LLM	Local LLM processes with per process CPU and RAM
🔥 HOT	Processes using >10% CPU (hidden when idle)
📊 MEM	Top memory consumers (>500 MB)
🌐 NET	Listening TCP ports (system ports filtered by default)

LLM detection

Automatically detects and tracks resource usage for:

Ollama, LM Studio, MLX LM, llama.cpp, llamafile, Claude, OpenAI CLI, Jan, GPT4All, KoboldCpp, Text Generation Inference, vLLM, LocalAI, Msty

Install

Quick (curl)

curl -fsSL https://raw.githubusercontent.com/elara-labs/syscheck/main/syscheck.sh -o /usr/local/bin/syscheck
chmod +x /usr/local/bin/syscheck

Or clone

git clone https://github.com/elara-labs/syscheck.git
cd syscheck
chmod +x syscheck.sh

Or just copy the script. It's a single file with no dependencies.

Usage

syscheck            # default (compact, system ports filtered)
syscheck --all      # show all ports, more memory entries
syscheck -a         # same as --all

Add to your shell

Drop this in your .zshrc or .bashrc to run it every time you open a terminal:

syscheck

Or alias it:

alias sc="syscheck"

How it works

Composes standard macOS tools (ps, lsof, df, top, vm_stat, ioreg, sysctl, pmset, netstat) into a single formatted dashboard. No sudo required, no background daemons, no temp files.

GPU utilization is read from the Apple Silicon AGX accelerator via ioreg, giving real device utilization, renderer utilization, and VRAM usage without elevated privileges.

Requirements

• macOS (Apple Silicon recommended for GPU metrics)
• Bash

License

MIT