Skip to main content
Send one daily digest of the changes that matter instead of an alert for every edit. A monitor judges each change against your goal, and your receiver stores events until the digest runs. You need STOPHY_API_KEY and a public address that accepts a POST. See the Quickstart.

1. Create the monitor

goal tells the judge which changes matter, and webhook says where events go. This request watches a page that returns a new identifier on every read, so a later check shows a change. The monthly schedule avoids daily checks while you build the receiver. Use { "text": "daily" } for a real digest, and replace the page and address with yours.
The answer holds the monitor’s id. Keep it in MONITOR_ID. The first check starts at once and sets the baseline. It sends monitor.check.completed and no monitor.page. See Targets.

2. Receive and store events

Verify the signature before you store anything. Get the account secret with POST /v1/webhook/secret and keep it in STOPHY_WEBHOOK_SECRET. This Bun receiver stores each verified event once, keyed by eventId, and answers before any digest work. The signature format is on Webhooks.
Run it with Bun.serve({ port: 3000, fetch: handle }) where the monitor can reach it.

3. See a change

After the baseline, run a check by hand instead of waiting for the schedule.
The check starts within a minute. Your receiver gets a monitor.page call like this one.
GET /v1/monitor/$MONITOR_ID/checks/$CHECK_ID returns the same result with the diff, whether or not your receiver was up. See Checks and changes.

4. Build the digest

Once a day, parse each stored body. Keep monitor.page events whose at is after your cutoff, and drop one only when judgment.meaningful is false, since a missing judgment does not rule out a change. Start the message with the number of changes. For each, list changeStatus, url and the judge’s reason, or No judgment. If nothing matches, send No changes worth reading in this window.

5. Catch what you missed

Before each digest, list failed deliveries and retry each id in data.deliveries with POST /v1/webhook/deliveries/{id}/retry. Your receiver ignores an eventId it already stored.
A page that could not be read is not a monitor.page event. Read the row’s error in GET /v1/monitor/{id}/changes. See Webhooks.

6. Clean up

An active monitor runs on its schedule. Delete the test monitor, then point a new one at your pages.

Check it

  • GET /v1/monitor/$MONITOR_ID/checks shows a completed first check.
  • A call with a wrong signature gets 401 and stores nothing. The same valid call twice stores one row.
  • The digest lists your changes and leaves out one the judge called noise.
  • GET /v1/webhook/deliveries?status=failed is empty on a normal day.
See Pricing for check costs, and Schedules to set how often a monitor checks.