Configuration Reference

When enabled, Pulse Endpoint automatically begins collecting metrics when the application launches. Useful for MDM deployments where monitoring should start without user interaction.

<key>autoStart</key>
<true/>

How often Pulse Endpoint collects and displays metrics (1-10 seconds). This unified setting applies to the popup refresh, local storage, and Splunk upload rate. Also configurable via CLI: pulse --start --interval 5

<key>intervalSeconds</key>
<integer>5</integer>

Maximum duration in minutes for a monitoring session. When set, Pulse Endpoint automatically stops collection after the specified time. 0 means unlimited (no auto-stop). Also configurable via CLI: pulse --start --duration 60. MDM takes precedence if set. CLI value is one-shot (cleared after monitoring starts).

<key>captureDurationMinutes</key>
<integer>60</integer>

List of process names to monitor separately with aggregated CPU, memory, network, and disk metrics. CLI-only (not MDM-managed). Set via repeatable --watch flags with --start. One-shot: cleared after monitoring begins.

pulse --start --watch "Google Chrome" --watch "zoom.us"

Show the menu bar icon but hide the detailed metrics preview in the popover. Only status indicators are shown.

<key>minimalUI</key>
<true/>

Controls log verbosity: 0 = Debug (all messages), 1 = Info, 2 = Warning (default), 3 = Error only. Logs are written to ~/Library/Logs/Pulse Endpoint/pulse.log. Set to 0 for troubleshooting, then restore to 2.

<key>logLevel</key>
<integer>0</integer>

Collect CPU load average (1m, 5m, 15m), usage breakdown (user, system, idle), and thermal throttle status. Essential for performance monitoring.

<key>collectCPU</key>
<false/>

Collect memory usage (active, wired, compressed, free), pressure level (normal, warning, critical), and swap activity. Critical for diagnosing slowdowns.

<key>collectMemory</key>
<false/>

Collect disk space usage (total, used, free) and I/O throughput (read/write KB/s). Useful for monitoring storage-intensive workloads.

<key>collectDisk</key>
<false/>

Collect network interface statistics (bytes/packets in/out, errors) and connectivity status. Useful for diagnosing connectivity and bandwidth issues.

<key>collectNetwork</key>
<false/>

Collect top CPU and memory consuming processes. This is the most resource-intensive metric as it spawns /bin/ps for system-wide process visibility. Disable to reduce overhead on resource-constrained systems.

<key>collectProcesses</key>
<false/>

Combine related processes by app name in the UI (e.g., all Chrome helpers shown as one entry). Useful when browsers/apps spawn many helper processes.

<key>aggregateProcesses</key>
<true/>

Collect GPU utilization percentage, memory usage, and top GPU processes. Keep enabled for creative/video workloads. Can disable for general office use.

<key>collectGPU</key>
<false/>

Collect battery state (charging, discharging, charged), power source (AC/battery), charge percentage, and sleep prevention assertions. Keep enabled for laptops. Can disable for desktop-only fleets.

<key>collectPower</key>
<false/>

Collect uptime, macOS version, hardware model, CPU brand, and chip description. Low overhead, provides essential inventory data for fleet management.

<key>collectSystem</key>
<false/>

Show process IDs in the process list UI. Disable for a cleaner display.

<key>showPIDsInUI</key>
<false/>

Store collected metrics locally in JSONL format. This is the default and recommended option.

<key>storageEnabled</key>
<true/>

Maximum disk space in megabytes for local metric storage. Oldest files are deleted when this limit is exceeded.

<key>maxStorageMB</key>
<integer>500</integer>

Number of days to retain metric files before automatic deletion.

<key>retentionDays</key>
<integer>30</integer>

How often to rotate metric files. Each rotation creates a new timestamped file.

<key>rotationPolicy</key>
<string>hourly</string>

Enable alert notifications. When disabled, no alerts are shown regardless of threshold violations.

<key>alertsEnabled</key>
<true/>

CPU usage percentage that triggers a high CPU alert.

<key>alertCPUThreshold</key>
<integer>80</integer>

Number of seconds CPU must exceed threshold before triggering alert. Set to 0 for immediate alerts.

<key>alertSustainedSeconds</key>
<integer>30</integer>

Memory pressure level that triggers alerts. 0=disabled, 2=warning and critical, 4=critical only.

<key>alertMemoryPressureLevel</key>
<integer>2</integer>

Minimum time between repeated alerts of the same type to prevent notification spam.

<key>alertCooldownMinutes</key>
<integer>30</integer>

Check thresholds even when monitoring is stopped. Uses 10-second check interval.

<key>alertInBackground</key>
<true/>

Seconds memory pressure must persist before alerting. 0 = immediate.

<key>alertMemorySustainedSeconds</key>
<integer>30</integer>

Show macOS notification when alert fires. Disable for silent logging only.

<key>alertShowNotification</key>
<false/>

Log alert events to storage for upload to Splunk. When disabled, alerts only trigger notifications.

<key>alertLogEvents</key>
<false/>

Custom title for CPU alert notifications. Leave empty for default ('High CPU Usage').

<key>alertCPUNotificationTitle</key>
<string>IT Alert: High CPU</string>

Custom body for CPU alert notifications. Placeholders: {value} = current CPU %, {threshold} = configured threshold.

<key>alertCPUNotificationBody</key>
<string>CPU is at {value}%, exceeds {threshold}%</string>

Custom title for memory alert notifications. Placeholder: {level} = 'warning' or 'critical'.

<key>alertMemoryNotificationTitle</key>
<string>Memory Alert ({level})</string>

Custom body for memory alert notifications. Placeholder: {level} = 'warning' or 'critical'.

<key>alertMemoryNotificationBody</key>
<string>Memory pressure is {level}. Please close unused apps.</string>

Base64-encoded RSA-signed license key. Can be deployed via MDM Configuration Profile or activated per-machine using the --license CLI flag.

<key>licenseKey</key>
<string>YOUR-LICENSE-KEY</string>