Skip to main content
Version: Next

System Info

This module provides information about the system such as:

  • CPU usage
  • Memory usage
  • Disk space
  • Network IP
  • Network speed
  • Temperature

It changes the indicator color based on the related value.
For example, if the CPU usage is above 80%, the indicator will be red.

By default, the module will display the CPU usage, memory usage, and temperature.

Indicators​

Using the indicators configuration, you can select which indicators to display in the status bar.

These are the available indicators:

CPU​

The CPU indicator displays the current CPU usage as a percentage by default.

To enable this indicator, add Cpu to the indicators configuration.

You can change the display format using the format option in [system_info.cpu]:

  • "Percentage" (default) — shows CPU usage as a percentage (e.g., 72%)
  • "Frequency" — shows the average CPU frequency in GHz (e.g., 3.42 GHz)

Memory​

The Memory indicator displays the current memory usage as a percentage by default.

To enable this indicator, add Memory to the indicators configuration.

You can change the display format using the format option in [system_info.memory]:

  • "Percentage" (default) — shows memory usage as a percentage (e.g., 64%)
  • "Amount" — shows used memory in GiB (e.g., 11.7 GiB)
  • "Fraction" — shows used and total memory in GiB (e.g., 5.12/15.89 GiB)

Memory Swap​

The Memory Swap indicator displays the current memory swap usage as a percentage by default.

To enable this indicator, add MemorySwap to the indicators configuration.

info

Memory Swap uses the same format setting as Memory ([system_info.memory]). If you set Memory to "Amount" or "Fraction", the swap indicator uses the same format.

Disk​

The Disk indicator displays the disk space usage for a specific path.

To enable this indicator, add { Disk = "path" } or { Disk = "path", Name = "label" } to the indicators configuration, where path is the mount point of the filesystem you want to monitor (e.g. / or /home, as shown by df) and label is an optional name to display for it.

warning

Disk matches the mount point, not the block device. A device node such as /dev/sda1 never matches, and the indicator is silently omitted from the bar.

You can change the display format using the format option in [system_info.disk]:

  • "Percentage" (default) — shows disk usage as a percentage (e.g., 54%)
  • "Fraction" — shows used and total disk space in GB (e.g., 256.00/512.00 GB)

Filtering Disk Mounts​

By default, all non-removable disks with available space are shown in the system info menu. You can use the mounts option in [system_info.disk] to specify which mount points to display.

If mounts is not specified, all disks are shown (default behavior).

[system_info.disk]
mounts = ["/", "/home"]
warn_threshold = 80
alert_threshold = 90
format = "Percentage"

With the above configuration, only the disks mounted at / and /home will be displayed in the menu.

Example​

To monitor the home directory disk space, you can add the following to your configuration:

[system_info]
indicators = [ { Disk = "/home" } ]

Or if you want to display the directory disk space with an optional name, for example bob instead of its full path:

[system_info]
indicators = [ { Disk = "/my/long/path/to/mount/called/bob", Name = "bob" } ]

IpAddress​

The IpAddress indicator displays the current IP address of the system.

To enable this indicator, add IpAddress to the indicators configuration.

DownloadSpeed​

The DownloadSpeed indicator displays the current download speed
of the system's network connection.

To enable this indicator, add DownloadSpeed to the indicators configuration.

UploadSpeed​

The UploadSpeed indicator displays the current upload speed
of the system's network connection.

To enable this indicator, add UploadSpeed to the indicators configuration.

Temperature​

The Temperature indicator displays the current temperature from the configured sensor.

To enable this indicator, add Temperature to the indicators configuration.

The sensor is configured with a single sensor option that accepts either a sensor type keyword (auto-detected) or an exact sensor label (manual override).

Sensor type keywords (auto-detected):

  • "Cpu" (default) — CPU temperature (coretemp/k10temp)
  • "Gpu" — GPU temperature (amdgpu/nouveau)
  • "Acpi" — ACPI thermal zone (acpitz)
  • "Nvme" — NVMe SSD temperature

Any other string is treated as an exact sensor label, overriding auto-detection (e.g. sensor = "k10temp Tctl"). If the configured label isn't found on your system, ashell falls back to Cpu auto-detection.

[system_info.temperature]
sensor = "Cpu"              # auto-detect by type
# sensor = "k10temp Tctl"   # or pin an exact sensor label

The temperature unit follows your locale / unit system (the global region option) by default. Set units to "Celsius" or "Fahrenheit" to override it for this indicator only — useful if you want Fahrenheit for the weather but Celsius for hardware temperatures.

[system_info.temperature]
units = "Celsius"           # override the locale unit system

ashell reads hardware sensors through the sysinfo crate, which builds each label from the hwmon data, so the value you write in sensor is not the bare label printed by sensors. The label is the chip name, followed by the sensor label (tempN_label) and the device model when they exist. A sensor with neither is named after its channel (temp1). For example:

  • k10temp Tctl: chip and sensor label
  • nvme Composite CT1000T705SSD3: chip, sensor label and device model
  • acpitz temp1: chip and channel, the sensor has no label

To list the labels exactly as ashell sees them, run this in sh or bash:

for d in /sys/class/hwmon/hwmon*; do
  chip=$(cat "$d/name")
  model=$(cat "$d/device/model" 2>/dev/null | sed 's/[[:space:]]*$//')
  for f in "$d"/temp*_input; do
    [ -e "$f" ] || continue
    n=$(basename "$f" _input)
    label=$(cat "$d/${n}_label" 2>/dev/null)
    [ -z "$label" ] && [ -z "$model" ] && label=$n
    echo "$chip${label:+ $label}${model:+ $model}"
  done
done

Alternatively, run ashell with level = "info" in the [logging] section: it logs the label of the sensor it auto-detected, and warns when a configured label was not found.

For NVMe SSDs, you'll need to find the model number first:

# Get NVMe model number
lsblk -d -o name,model
# Output example:
# NAME    MODEL
# nvme0n1 CT1000T705SSD3

Common sensor labels include:

  • acpitz temp1 - ACPI thermal zone
  • coretemp Package id 0 - Intel CPU temperature
  • k10temp Tctl / k10temp Tdie - AMD Ryzen CPU temperature
  • amdgpu edge - AMD GPU temperature
  • nvme Composite MODEL_NAME - NVMe SSD temperature (use model from lsblk output)

Polling Interval​

You can configure how often the system information is refreshed using the interval option (in seconds). The default is 5 seconds.

[system_info]
indicators = [ "Cpu", "Memory", "Temperature" ]
interval = 10

Higher values reduce CPU usage at the cost of less frequent updates.

Display Formats​

The Cpu, Memory and Disk indicator types each support a format option that controls how their value is displayed in the status bar and menu. The format is configured in the corresponding [system_info.<type>] section.

MemorySwap has no section of its own: it reuses [system_info.memory] for both its format and its thresholds.

The remaining indicators (Temperature, IpAddress, DownloadSpeed, UploadSpeed) always show their value directly and have no format option.

info

Warning and alert color thresholds remain active regardless of the display format. For temperature, thresholds are interpreted in the displayed unit (units, or the one determined by your locale / unit system), so if the indicator shows Fahrenheit, set your thresholds in Fahrenheit. The built-in temperature defaults are 60 / 80 in Celsius and 140 / 176 in Fahrenheit; ashell picks the pair matching the resolved unit.

Example​

[system_info.cpu]
format = "Frequency"

[system_info.memory]
format = "Fraction"

[system_info.disk]
format = "Fraction"

Warning and Alert Thresholds​

You can also configure the warning and alert thresholds for the following indicators:

  • CPU
  • Memory (RAM and Swap use the same thresholds)
  • Disk
  • Temperature

Each one is configured in its own table, [system_info.<indicator>], where <indicator> is one of cpu, memory, disk or temperature:

[system_info.cpu]
warn_threshold = 60
alert_threshold = 80
info

warn_threshold must be lower than alert_threshold. If it is not, ashell logs a warning and lowers warn_threshold to match alert_threshold.

MemorySwap has no table of its own and reuses the [system_info.memory] thresholds.

Default Configuration​

[system_info]
indicators = [ "Cpu", "Memory", "Temperature" ]
interval = 5

[system_info.cpu]
warn_threshold = 60
alert_threshold = 80
format = "Percentage"

[system_info.memory]
warn_threshold = 70
alert_threshold = 85
format = "Percentage"

[system_info.disk]
warn_threshold = 80
alert_threshold = 90
format = "Percentage"
# mounts = ["/", "/home"]  # uncomment to whitelist specific mount points

[system_info.temperature]
sensor = "Cpu"  # type keyword ("Cpu", "Gpu", "Acpi", "Nvme") or an exact sensor label
# sensor = "k10temp Tctl"  # example: pin an exact sensor label
# warn_threshold and alert_threshold are unset by default. Omit them and ashell
# uses 60 / 80 in Celsius, or 140 / 176 when the resolved unit is Fahrenheit.
# units is unset by default: omit it to follow the locale unit system.
warning

TOML has no None literal. Leave warn_threshold, alert_threshold and units out of the file to keep their defaults. Writing warn_threshold = None is a parse error that makes ashell fall back to the entire default config.