Configuration

Monitor reads its settings from a single JSON file:

~/.config/monitor/config.json

The file is optional. If it does not exist — or if it contains invalid JSON — Monitor falls back to the built-in defaults rather than failing. The directory (~/.config/monitor) is created automatically when settings are saved.

Settings drive the TUI (update cadence, process limits, mouse, alert thresholds). monitor processes also uses max_processes and show_system_processes as its defaults; explicit CLI flags still win.

Manage settings from the CLI

The config command avoids hand-editing duration values and validates every change before atomically replacing the file:

monitor config show
monitor config get update-interval
monitor config set update-interval 500ms
monitor config set cpu-alert-threshold 85
monitor config path
monitor config reset --yes

Every subcommand also supports --json. Keys accept either underscores or hyphens. config set accepts Go-style durations such as 250ms, 2s, and 1m, then stores them in the existing nanosecond representation.

Example

A config file with every field set to its default:

{
  "update_interval": 1000000000,
  "temperature_unit": "C",
  "show_system_processes": false,
  "max_processes": 50,
  "mouse_enabled": true,
  "cpu_alert_threshold": 0,
  "memory_alert_threshold": 0
}

You only need to include the fields you want to override — any field you omit keeps its default.

Fields

Field	Type	Default	Description
update_interval	duration (nanoseconds)	1000000000 (1s)	How often the collector refreshes metrics. Encoded as an integer number of nanoseconds (Go time.Duration). Values <= 0 are ignored; full process collection clamps positive values below 100ms to 100ms.
temperature_unit	string	"C"	Temperature display unit (C or F, case-insensitive). Unknown values are ignored and the default is kept.
show_system_processes	bool	false	Whether system/root-owned processes are shown in the Processes tab.
max_processes	int	50	Maximum number of processes listed. Only accepted in the range 1–1000; out-of-range values are ignored and the default is kept.
mouse_enabled	bool	true	Whether mouse interaction (clicking tabs, selecting rows, scrolling) is enabled in the TUI.
cpu_alert_threshold	float	0	CPU-usage percentage that triggers an alert. 0 disables the alert. Values outside 0–100 are ignored.
memory_alert_threshold	float	0	Memory-usage percentage that triggers an alert. 0 disables the alert. Values outside 0–100 are ignored.

update_interval is in nanoseconds

Because update_interval is a Go time.Duration, it is stored as an integer number of nanoseconds, not a string. For example, 500000000 is 500ms and 2000000000 is 2 seconds.

How the file is written

When Monitor saves settings it writes them atomically: the JSON is first written to a temporary file in the same directory, then renamed into place. rename is atomic, so a crash or a full disk mid-write cannot leave a truncated config.json that would otherwise be silently discarded in favour of the defaults. The resulting file is written with mode 0644.

Invalid or missing config

Monitor is deliberately forgiving here:

• Missing file → defaults are used.
• Unreadable / malformed JSON → defaults are used (the file is not rewritten or deleted).
• Out-of-range field (e.g. max_processes of 0 or 5000, a threshold outside 0–100, an unknown temperature unit, or a non-positive update_interval) → that single field falls back to its default; the rest of the file is still applied.

Temperature source flag

Separate from the config file, a persistent CLI flag controls how temperature is read. It is a persistent flag, so it is available on every subcommand, including the TUI (monitor studio):

monitor studio --no-temperature-source
# or for a one-shot:
monitor snapshot --no-temperature-source

By default Monitor wires a powermetrics-backed temperature source so readings come from real SMC sensors when sudo credentials are available, falling back to a CPU-load estimate otherwise. Passing --no-temperature-source skips the powermetrics subprocess entirely and always uses the estimation heuristic — useful on systems where sudo can't be obtained, with no recompile required.

Environment variables

monitor run sets two environment variables in the glyphrun child so the spec and the processes it launches can detect the observed run:

Variable	Set by	Meaning
MONITOR	monitor run	Always set to 1 in the glyphrun child, mirroring glyphrun's GLYPHRUN pattern.
MONITOR_RUN_DIR	monitor run	Reuses a non-empty inherited value; otherwise points to a temporary directory that Monitor removes after glyphrun exits.
MONITOR_LOG_STORE	monitor logs capture/search	Shared override for the log database path. A command-level --store flag takes precedence; the default is ~/.local/share/monitor/logs.veclite.
MONITOR_ISSUES_STORE	issue CLI, investigate, watch, MCP	Shared override for the issue database path. issues --store takes precedence for CLI reads/mutations; the default follows $XDG_DATA_HOME/monitor/issues.veclite or ~/.local/share/monitor/issues.veclite.

A glyphrun child can check MONITOR=1 (or the presence of MONITOR_RUN_DIR) to alter its behaviour — for example to emit extra diagnostics — while being monitored.

Diagnostic run context

investigate and watch --stash also read optional correlation fields. An explicit investigate flag wins over MONITOR_*, which wins over Chalupa CI and legacy aliases.

Field	Preferred environment variables
environment	MONITOR_ENVIRONMENT, CHALUPA_CI_ENVIRONMENT, CHALUPA_ENV, CHALUPA_ENVIRONMENT
deployment	MONITOR_DEPLOYMENT_ID, CHALUPA_DEPLOYMENT_ID
run	MONITOR_RUN_ID, CHALUPA_CI_RUN_ID, CHALUPA_RUN_ID, basename of MONITOR_RUN_DIR
step	MONITOR_STEP_ID, CHALUPA_CI_STEP_ID
suite	MONITOR_SUITE, CHALUPA_CI_SUITE
attempt	MONITOR_ATTEMPT, CHALUPA_CI_ATTEMPT
release	MONITOR_RELEASE, CHALUPA_RELEASE
service	MONITOR_SERVICE, CHALUPA_SERVICE
Git SHA	MONITOR_GIT_SHA, GIT_SHA, GITHUB_SHA

These values belong to the local diagnostic plane and occurrence context. They are never added to monitor.telemetry_window V1.