MacOS Live Monitor & Diagnostics

A Bash-based macOS system monitoring dashboard that runs directly in your terminal. It shows a health check (with OK / WARNING / CRITICAL statuses) and detailed resource metrics on a single screen, using a minimalist monochrome style with a live spinner animation.

All logic lives in one file: monitor.sh. No external dependencies.

---

Features

1. System Diagnostics & Health Check

Each component is labelled [✓] OK, [!] WARNING, [✕] CRITICAL, or [i] INFO.

Component	Data source	Thresholds
Disk Usage	df -h /	≥ 85% → CRITICAL, ≥ 75% → WARNING
Disk S.M.A.R.T	diskutil info /	anything other than Verified → CRITICAL (possible SSD failure). The line is hidden when the status cannot be read
Memory & Swap	sysctl vm.swapusage + memory_pressure	swap > 2048 MB → WARNING, or memory pressure reporting warn/critical → WARNING
CPU / Runaway Process	ps -eo %cpu,pid,comm -r	top process at ≥ 80% → CRITICAL, ≥ 50% → WARNING (shows process name + PID)
Battery Health	pmset -g batt	Service Battery / Replace Soon → CRITICAL; battery ≤ 15% while unplugged → CRITICAL; desktop without a battery → INFO

2. Detailed Resource Metrics

Three pie charts rendered side by side directly in the terminal, drawn with half-block characters (▀ / ▄) so each pie is a 12×12 pixel circle. The three columns plus their gaps measure exactly 42 columns, lining up with the dividers used elsewhere in the dashboard.

Pie	Slices	Source
CPU Usage	User / Sys / Idle	top -l 1
Memory	Used / Free	memory_pressure free percentage
Storage	Used / Free	df -h /

Colour rule — a pie is green while healthy and turns red once utilisation reaches 90% (User + Sys for CPU, Used for memory and storage). The second slice of the CPU pie (Sys) uses a lighter step of the same hue, and the unused share always stays a recessive grey. The threshold lives in the ALERT_AT variable near the top of the script — change it in one place to move the alert point.

Every pie carries a legend with the numeric percentage and a headline value marked [✓] or [✕], so the alert state is never signalled by colour alone (readable for colourblind users, on monochrome terminals, and when piped through sed).

Other metrics in this section:

• Memory Pressure — condensed memory_pressure output, printed below the pies
• Storage detail — used, total, and free space in human-readable units
• Battery Status — charge level and state (charging / discharging / AC attached); this section is hidden automatically on machines without an internal battery
• System Uptime — how long the system has been running

Battery is deliberately not drawn as a pie and is excluded from the 90% rule: a battery at 95% is a healthy state, not an alert, so the same threshold would invert the meaning.

3. Display Quality

• Flicker-free — every metric is collected first, then the screen is redrawn once per cycle, so there is no flashing or tearing
• Live spinner & spark star — ⠋⠙⠹… and ✦✧✶✵ animations signal that the dashboard is alive
• Cursor hiding — the cursor is hidden (tput civis) during live mode for a clean view
• Graceful exit — a trap on Ctrl+C / SIGTERM restores the cursor (tput cnorm) and prints a closing message, so the terminal is never left in a broken state

---

Requirements

• macOS (the script uses diskutil, pmset, memory_pressure, and sysctl vm.swapusage)
• Bash and awk (both ship with macOS by default)
• A terminal that supports 256 colours, ANSI escape codes, and Unicode block characters — Terminal.app, iTerm2, Ghostty, and VS Code's terminal all qualify

---

Usage

Make the script executable once:

chmod +x monitor.sh

Live mode (default)

Auto-refreshes every 1 second:

./monitor.sh

Live mode with a custom interval

The first argument is the refresh interval in seconds:

./monitor.sh 5     # refresh every 5 seconds
./monitor.sh 0.5   # refresh twice per second

Single-run mode (snapshot)

Print the dashboard once and exit — useful for logging or calling from another script:

./monitor.sh --once
# or
./monitor.sh -1

Exiting live mode

Press Ctrl+C. The terminal cursor is restored automatically.

Changing the alert threshold

Edit ALERT_AT near the top of monitor.sh — it controls the utilisation percentage at which a pie flips from green to red:

ALERT_AT=90     # default: red at 90% and above

---

Sample Output

┌──────────────────────────────────────────┐
│   MacOS Live Monitor & Diagnostics  [⠹] │
└──────────────────────────────────────────┘

❖ System Diagnostics & Health Check
──────────────────────────────────────────
[✓] OK: Disk capacity is healthy (26% used).
[✓] OK: SSD/Disk S.M.A.R.T Status Verified.
[!] WARNING: High swap usage (2567 MB used).
[!] WARNING: 'Antigravity IDE Helper (Renderer)' (PID: 29231) is using 71% CPU.
[✓] OK: Battery is normal (69% - charging).

❖ Detailed Resource Metrics
──────────────────────────────────────────

◇ CPU Usage    ◇ Memory       ◇ Storage
  ▄▄▀▀▀▀▄▄       ▄▄▀▀▀▀▄▄       ▄▄▀▀▀▀▄▄
 ▀▀▀▀▀▀▀▀▀▀     ▀▀▀▀▀▀▀▀▀▀     ▀▀▀▀▀▀▀▀▀▀
▀▀▀▀▀▀▀▀▀▀▀▀   ▀▀▀▀▀▀▀▀▀▀▀▀   ▀▀▀▀▀▀▀▀▀▀▀▀
▀▀▀▀▀▀▀▀▀▀▀▀   ▀▀▀▀▀▀▀▀▀▀▀▀   ▀▀▀▀▀▀▀▀▀▀▀▀
 ▀▀▀▀▀▀▀▀▀▀     ▀▀▀▀▀▀▀▀▀▀     ▀▀▀▀▀▀▀▀▀▀
  ▀▀▀▀▀▀▀▀       ▀▀▀▀▀▀▀▀       ▀▀▀▀▀▀▀▀
● User 12.8%   ● Used 65.0%   ● Used 26.0%
● Sys  18.7%   ○ Free 35.0%   ○ Free 74.0%
○ Idle 68.5%
[✓]   31.5%    [✓]   65.0%    [✓]   26.0%

  System-wide memory free percentage: 35%
  Storage (/): Used 12Gi / Total 228Gi | Free 35Gi

◇ Battery Status
  Level: ✦ 69% (charging)

◇ System Uptime
  Uptime:  4:18  6 users

──────────────────────────────────────────
  [ Press Ctrl+C to exit | Live ⠹ | Auto-refreshing every 1s ]

The pie slices above are green in a real terminal; the colour is lost here because the sample was captured with ANSI codes stripped. Once a pie crosses ALERT_AT its slices turn red and its headline marker changes from [✓] to [✕].

In --once mode the title becomes MacOS System Monitor & Diagnostics (no spinner) and the auto-refresh footer line is not printed.

---

Notes

• The interval is read from the first argument, so --once / -1 cannot be combined with an interval value. Both flags disable the loop entirely, so the interval is never used there.
• Very small intervals (e.g. 0.2) call top -l 1 and diskutil very frequently, which adds CPU load of its own. An interval of 1–5 seconds is a reasonable compromise.
• diskutil and pmset calls are wrapped with 2>/dev/null, so the dashboard still renders even if those commands fail or are unavailable.
• The System Uptime line also includes the number of logged-in users (e.g. 3:55 6 users) because it takes the first two fields of the uptime output.