A floating macOS system monitor — CPU, GPU, memory, network, disk, and Ollama model status in a compact always-on-top panel.
- Zero config — launches with sensible defaults, auto-creates
~/.config/statbar/config.json - Hot-reload — edit
config.jsonwhile running, changes apply instantly - Live stats — CPU (Mach), GPU (IOKit/AGXAccelerator), memory (Mach VM), network (
getifaddrs), disk (APFS volume) - Ollama integration — shows running models with parameter size, quantization, context length, and GPU usage
- Configurable layout — 6 stat sections in any order, each individually toggleable
- MCP server — configure remotely via Hermes Agent or any MCP-compatible client
- Pure Swift + Python stdlib — zero dependencies beyond macOS 14+
git clone https://github.com/stuartmhannon/statbar.git
cd statbar
./compile.sh
open StatBar.appAdd StatBar.app to Login Items for auto-launch on boot.
All settings live in ~/.config/statbar/config.json. The file auto-creates on first launch.
Key sections:
| Section | What it controls |
|---|---|
window |
Position, size, min/max height |
appearance |
Background color/opacity, shadow, titlebar |
stats |
Visibility, colors, section order, network interface prefixes |
ollama |
Endpoint URL, timeout, enabled/disabled |
When statbar-mcp is running, Hermes Agent can control StatBar remotely:
statbar_get_config— dump current configstatbar_update_config— partial deep-merge (change any subset of keys)statbar_reset_config— restore defaultsstatbar_get_status— check if StatBar is running
- macOS 14.0+
- Apple Silicon (optional, Intel works for CPU/memory/network/disk; GPU requires AGXAccelerator)
- Command Line Tools (
swiftc,python3)
MIT
