A local backup alerting tool that watches PulsePoint Respond for Web and alerts when a configured unit appears.
Backup alert only. This is not official dispatch, not a pager, not CAD, not a radio, and not a life-safety system. Do not rely on it as your sole means of receiving emergency calls.
v0.1.0-alpha.1 — working prototype being refactored into a cross-platform app.
- Monitor one or more PulsePoint agency feeds.
- Watch one or more units/apparatus IDs.
- Choose Alert Me or Track Unit(s) behavior for monitored-unit activity.
- Test mode: alert on any new PulsePoint page activity.
- Laptop audible alert.
- Alert until acknowledged.
- Windows sleep prevention while monitor runs.
- Pushover emergency push support.
- ntfy urgent push support.
- Local web UI at
http://127.0.0.1:8765. - Saved agency and unit presets.
- First-run setup wizard.
- Installer copies the default alert sound into the runtime config folder.
- App version is shown in the UI.
From a cloned copy of this repository:
cd C:\Users\Shane\Documents\GitHub\pulsepoint-alerts
.\installers\windows\install.bat
.\installers\windows\start.batThen open:
http://127.0.0.1:8765
The Windows installer prepares the runtime folder:
C:\pulsepoint-alert
and copies the default alert sound to:
C:\pulsepoint-alert\alert.wav
The app includes a first-run setup wizard at:
http://127.0.0.1:8765/first-run
Use it to configure the minimum required settings:
- Agency/feed name
- PulsePoint agency ID(s)
- Unit set name
- Unit IDs to monitor
- Poll interval
- Test mode
- Sleep prevention
After saving the wizard, configure laptop and phone alerts from the Alerts page.
- Start the app.
- Open the local web interface.
- Configure agencies and unit sets.
- Configure laptop and phone alerts.
- Use test mode to confirm the alert path.
- Turn test mode off for normal unit-watch mode.
- Start the monitor.
python -m venv .venv
source .venv/bin/activate
python -m pip install -e .[dev]
python -m playwright install chromium
python -m pulsepoint_alertsWindows PowerShell:
python -m venv .venv
.\.venv\Scripts\Activate.ps1
python -m pip install -e .[dev]
python -m playwright install chromium
python -m pulsepoint_alertsSee docs/.
Apache License 2.0. See LICENSE and NOTICE.
See DISCLAIMER.md.
PulsePoint Alert Monitor is a supplemental awareness and backup alerting aid only.
This project is not affiliated with, endorsed by, sponsored by, or approved by PulsePoint Foundation, PulsePoint Respond, or any public safety agency. “PulsePoint” and related names may be trademarks or service marks of their respective owners.
This software is not an official dispatch system, pager, CAD terminal, MDT, radio, station alerting system, or life-safety system. Do not rely on it as your sole or primary means of receiving emergency calls or dispatch information.
PulsePoint feeds may lag, omit incidents, omit units, change format, become unavailable, or fail. This tool may also fail due to operating system sleep, network connectivity, browser changes, API/provider failures, local configuration errors, notification-service failures, device settings, or software defects.
This software is provided “as is,” without warranty of any kind. The authors, maintainers, contributors, and affiliated entities are not responsible for missed calls, delayed responses, operational errors, injury, death, data loss, device malfunction, notification failure, employment consequences, disciplinary action, or any other damages or bad outcomes arising from use or misuse of this software.
The Alerts page includes a setting to control whether call details are included in phone push notifications.
When enabled, Pushover and/or ntfy messages may include incident details from the PulsePoint Active section. Depending on phone settings, these details may appear on the lock screen.
Turn this setting off if you want phone pushes to show only that a monitored unit alert occurred without including call details.
Choose the active profile on the Alerts page:
- Alert Me keeps the full alert workflow: enabled laptop/desktop looping alerts, emergency or repeating phone pushes, ACK/Silence requirements, history, and evidence.
- Track Unit(s) records monitored-unit activity without starting the looping desktop alert. Enabled phone pushes use low-priority, non-emergency delivery and do not request ntfy calling. Tracking events require no ACK while still retaining history and evidence.
The selected profile is shown on the Alerts page, Dashboard, top status bar, History, Evidence, Troubleshooting, diagnostics, and exported data. Existing configurations default to Alert Me.
Alert history is saved locally in the app runtime folder as alert_history.json.
Default runtime locations:
- Windows:
C:\pulsepoint-alert\alert_history.json - macOS/Linux:
~/.pulsepoint-alerts/alert_history.json
Alert history survives app restarts, can be cleared from the History page, and can also be exported as CSV.
Manual laptop alert tests, manual phone push tests, and simulated active incident alerts are written to Alert History.
Phone push test entries are marked as manual_phone and as not requiring acknowledgement because they do not create a local laptop alert that needs to be silenced.
PulsePoint Alert Monitor includes a Windows shortcut helper script.
The Troubleshooting page also provides cross-platform Enable Start at Login and Disable Start at Login controls. These use the current user's Windows Startup folder, macOS LaunchAgents, or Linux XDG autostart directory. No administrator access is required.
From the repository folder, run:
powershell -ExecutionPolicy Bypass -File .\installers\windows\create_shortcuts.ps1
This creates a desktop shortcut named:
PulsePoint Alert Monitor
The shortcut launches the app through:
installers\windows\start.bat
To launch the app automatically when Windows starts, run:
powershell -ExecutionPolicy Bypass -File .\installers\windows\create_shortcuts.ps1 -StartAtLogin -NoDesktop
This creates a shortcut in the Windows Startup folder.
Run:
C:\Users\Shane\AppData\Roaming\Microsoft\Windows\Start Menu\Programs\Startup\PulsePoint Alert Monitor.lnk = Join-Path ([Environment]::GetFolderPath("Startup")) "PulsePoint Alert Monitor.lnk"
Test-Path C:\Users\Shane\AppData\Roaming\Microsoft\Windows\Start Menu\Programs\Startup\PulsePoint Alert Monitor.lnk
Expected result:
True
Run:
powershell -ExecutionPolicy Bypass -File .\installers\windows\create_shortcuts.ps1 -RemoveStartAtLogin -NoDesktop
Start-at-login launches the app. To make the monitor itself start automatically after the app launches, enable the in-app setting:
Auto-start monitor when app launches
This setting is located on the Monitor Setup page.
Developers can install a local pre-commit documentation guard:
powershell -ExecutionPolicy Bypass -File .\scripts\install_git_hooks.ps1
After installation, commits that change source, installer, or test files will require a related documentation or roadmap update.
This helps keep README.md, ROADMAP.md, CHANGELOG.md, CONTRIBUTING.md, and docs/ aligned with code changes.
The Config page provides local configuration management.
Available actions:
- Export Full Config JSON
- Export Redacted Config JSON
- Import Config JSON
- Reset Config to Defaults
- View the active config file path
Full config exports may include sensitive values such as Pushover keys or ntfy tokens. Store full exports securely.
Use redacted exports for troubleshooting or sharing because token/key fields are replaced with REDACTED.
Resetting config does not delete alert history.
The Config page includes an Export Diagnostics ZIP button.
The diagnostics ZIP includes:
- diagnostics.json
- redacted_config.json
- recent_logs.txt
- alert_history_recent.json
The diagnostics export is intended for troubleshooting. It uses a redacted config file so Pushover keys and ntfy tokens are not included.
The full config export is still available separately for personal backup and restore, but full exports may contain secrets and should be stored securely.
The test suite includes parser-safety tests for the PulsePoint Active/Recent split.
These tests verify that:
- Active incidents are scanned
- Recent/closed incidents are ignored
- monitored unit IDs are detected in Active incidents
- time/age noise is normalized out of incident signatures
- repeated lines are deduplicated in call-detail summaries
The live monitor uses domcontentloaded plus a short page-settle delay instead of waiting for full network idle, because PulsePoint can continue background network activity and cause network-idle waits to hang.
The top status bar displays the active PulsePoint agency ID(s). When the active agency ID string matches a saved agency preset, the saved agency name is displayed beside the ID(s).
Example:
Agency IDs: 37047 (AMR San Diego)
If no saved agency preset matches the active ID string, the status bar displays the raw agency ID(s).
The dashboard, config summary, and top status bar display saved preset names when the active agency ID(s) or unit list match a saved preset.
Examples:
Agency: 37047 (AMR San Diego)
Units: E36, M231 (Station 36 / Jamul set)
If no saved preset matches, the app displays the raw agency ID(s) or unit list.
The dashboard and top status bar include monitor-health fields so the user can tell whether the monitor is actively checking PulsePoint.
Displayed health fields include:
- Last check time
- Last successful check time
- Last page refresh time
- Consecutive error count
- Last error
- Whether the Active section was found during the latest successful unit-mode scan
These fields are also included in the diagnostics ZIP.
The monitor-health display uses human-readable age text and color-coded health badges.
Health states:
- STOPPED: monitor is not running
- WAITING: monitor is running but no successful check has completed yet
- HEALTHY: recent successful check and no consecutive errors
- DEGRADED: one or two consecutive errors
- ERROR: three or more consecutive errors
- STALE: last successful check is older than the stale threshold
The stale threshold is calculated from the configured poll interval.
The Dashboard auto-refreshes every 10 seconds so monitor health, last-check time, error count, and status badges stay current without manual browser refresh.
Configuration pages do not auto-refresh. This prevents losing typed settings while editing agencies, units, alerts, or config values.
The Dashboard includes a Refresh PulsePoint Now button.
This button requests an immediate reload of the PulsePoint page on the next monitor cycle. It does not restart the monitor and does not change configuration.
The button is disabled when the monitor is stopped.
The top status bar is intentionally kept as an at-a-glance summary:
- Monitor
- Health
- Alert
- Agency
- Units
Detailed health information is shown in the Dashboard Monitor Health card to avoid duplicating too much information across the page.
The monitor splits the PulsePoint Active section into incident-sized blocks before looking for monitored units.
This prevents a monitored unit on one incident from being bundled with unrelated neighboring incidents. Earlier sliding-window parsing could cause false alerts when a new nearby incident appeared above or below an existing monitored-unit incident.
Incident signatures ignore unit-only lines so responder-list changes, unit-status marker changes, or unit ordering changes do not create duplicate alerts for the same incident.
If incident boundaries cannot be identified, the parser fails safe rather than bundling the entire Active section into one incident.
Real monitor alerts save a local evidence snapshot at the moment of alert.
Evidence snapshots include the matched monitored units, new incident signatures, triggering incident block text, configured units, agency IDs, signature method, and the raw PulsePoint Active section from that check cycle.
Alert History links to evidence snapshots when available. Diagnostics ZIP exports include recent alert evidence snapshots as alert_evidence_recent.json.
Evidence may contain call details or addresses from PulsePoint Active and should be treated as sensitive troubleshooting data.
The app includes a read-only Troubleshooting page showing app version, Python/platform information, Playwright package status, runtime paths, shortcut status, monitor health, configured units, alert history count, evidence snapshot count, and quick export buttons for diagnostics, alert history, and redacted config.
The RUNNING / STOPPED monitor pill in the top status bar is clickable. Clicking it starts or stops the monitor without returning to the Dashboard.
The app header also displays the PNG app icon inside the program UI. This is separate from the favicon and Windows shortcut .ico.
Current alpha version: 0.2.0-alpha.1
See CHANGELOG.md for release notes.
When running from a source checkout on Windows, the Start at Login control uses installers/windows/start.bat so startup launches the app through the same tested path used for manual startup. If the app is packaged or installed differently, it falls back to python -m pulsepoint_alerts.
When the monitor cannot find the PulsePoint Active section, it fails safe and does not scan the whole page. To support troubleshooting, PulsePointer Alerter records one full-page text snapshot per missing-Active streak under the local runtime folder:
debug_snapshots/
These files are local diagnostic artifacts and may contain PulsePoint page text, call details, addresses, and unit information. Do not post them publicly unless reviewed/redacted.
A public PDF user guide is available in:
docs/PulsePointer_Alerter_User_Guide_v0.2.0-alpha.1.pdf
The editable DOCX source is kept locally and is not required for normal users.
Troubleshooting can list recent debug snapshots created when the PulsePoint Active section is unavailable. Snapshot files are local diagnostic artifacts and may contain call/location details.
PulsePointer Alerter supports multiple phone alert options so users can choose the right cost/reliability tradeoff.
For the most complete implementation and best wake-up behavior, Pushover is the recommended provider. Pushover emergency-priority alerts can repeat until acknowledged, and Pushover receipt support makes it the best future path for phone-side acknowledgement tracking and stopping a laptop alert after a phone acknowledgement.
Recommended options:
- Free/default path: ntfy
- Best full-feature path: Pushover
- Maximum redundancy: Both ntfy and Pushover
- Future paid fallback: Twilio/SMS
Pushover is not required to use the app, but it is the best bargain for the full overnight wake-up and acknowledgement workflow.
If the monitored unit list is changed while the monitor is running, PulsePointer Alerter recaptures the unit baseline before alerting. This prevents already-active calls from alerting simply because a newly added unit now matches the current Active list.
The Dashboard ACK/Silence control is disabled and grey when no alert is active. When an alert is active, the button becomes enabled and flashes red to draw attention.
When Pushover emergency-priority alerts return a receipt, PulsePointer Alerter stores that receipt with the alert history record. Desktop ACK/Silence cancels further Pushover emergency retries for that receipt. If the notification is acknowledged in Pushover, the app polls the receipt status and mirrors that phone acknowledgement into local alert history while silencing the laptop alert.
This ACK bridge is Pushover-specific. ntfy alerts do not currently provide a reliable local acknowledgement signal back to the app.
Alert Me mode is latency-sensitive. In Alert Me mode, PulsePointer refreshes the PulsePoint page before each configured poll scan so the monitor is not repeatedly scanning stale browser text between slower full-page refreshes. Track Unit(s) may continue to use the configured slower refresh interval to reduce churn.
If the Active section is temporarily unavailable, the app skips that cycle, preserves the existing active-call baseline, and does not scan Recent/closed calls.
For Pushover emergency-priority alerts, acknowledging the alert on the phone is mirrored into local alert history and also force-stops any active desktop/laptop audible alert. Desktop ACK/Silence also cancels further Pushover emergency retries when a receipt is available.
Pushover ACK polling uses separate state from the desktop alert sound loop. This prevents a stale desktop ACK stop event from blocking Pushover receipt polling or follow-up Alert Me alerts.
Core source files include module docstrings and function-level comments for the alert pipeline, monitor decision logic, parser safety rules, acknowledgement flow, and local runtime persistence. Safety-sensitive comments are intended to explain why the code behaves conservatively, especially around Active-vs-Recent separation, baseline handling, test alert isolation, and cross-device ACK behavior.
PulsePointer Alerter treats both same-day time lines, such as 9:23 AM, and relative time lines, such as Yesterday 5:09 PM, as Active incident boundaries. This prevents long-running incidents from being merged into neighboring monitored-unit calls when the Active list changes size.