About HookPulse

HookPulse answers one question about a routine that is supposed to run on its own: did it run, or did it go silent? This page says how the check works, what it costs and what it is not.

A dead-man switch, not an uptime checker

An uptime checker fetches your site and complains when it fails. That cannot see a cron job that stopped being scheduled, a backup that never finished, an n8n scenario that was paused, or a webhook handler that silently stopped processing. HookPulse turns the direction around: each monitor has a private check-in URL, your routine calls it when it finishes, and silence for longer than the interval you chose is the signal.

The interval goes from 90 seconds to 24 hours. Any HTTP method counts as a ping, and every ping is stored as an event with its time and status, so you can see when the routine last ran.

How the check runs

Every 5 minutes, a scheduled job on our side compares each monitor's last ping with its interval. Past the interval, the monitor is marked overdue and a miss is recorded: we send the e-mail alert and POST a JSON body to the webhook URL you gave — the exact body is returned by GET /api/templates for the monitor, so nothing has to be guessed. Alerts are de-duplicated: at most one per 24 hours, or per interval if the interval is longer. The next successful ping clears the state.

Each owner also has a public status page as JSON and RSS, so a team or an agent can read the state of all monitors without a token.

What is free and what costs

What HookPulse is not

Who makes it

HookPulse is one of the house's products, built to be used by people and by AI agents through the same API. Questions, bugs or requests: contato@hookpulse.net.

Back to the monitors