Your site can look fine
and still be broken.

Yes. Paste any address into the box on the auguro.dev home page and it runs a real browser check there and then, showing you the verdict, what it found, and a screenshot of what our browser saw. No account and no card. Signing up is what turns that one-off look into something that keeps watching and tells you when it changes.

That the page rendered. Auguro loads your site in real Chromium and checks that there is content on the page, that your own scripts did not throw, that your own stylesheets and scripts loaded, and that whatever you told it to look for is still there. A server can answer 200 with a completely blank page, and that is invisible to a monitor that only reads the response code.

No. You paste a URL. If you want to be stricter you can add text that must appear, text that must not appear, a CSS selector that has to exist, and a response time ceiling, but none of it is required and none of it is code.

By only failing a check on things that are the site’s own fault. A third-party advertising script throwing an error does not fail the check. A lazy-loaded image below the fold is not a broken image. A missing favicon or a telemetry beacon returning 400 is a warning, not an outage. Anything ambiguous warns and never pages you.

Fifteen seconds by default, and you can set anything from five to thirty seconds per monitor. The clock starts the moment we make the request to your site and stops when we have the answer.

Whatever you set. Every check has a threshold for how many consecutive failures count as an outage and how many consecutive successes count as recovered, so a single blip never wakes anyone.

In the EU. Every organization’s data lives at Cloudflare pinned to the EU jurisdiction. The same for screenshots and saved page sources. Initial checks also run from the EU by default. The one thing that changes that is choosing extra regions yourself: a monitor sampled from Sydney is genuinely run from Sydney, so the request to your site starts there and the response passes through that data centre on its way to storage in the EU.

No, and we will not. Billing is the number of sites and how often we look at them, nothing else. Charging per teammate punishes you for adding the person who would actually answer the alert.

Slack, Telegram, email, Pushover and webhooks. One incident groups every notification about a single outage, so a flapping check cannot generate a stream of duplicate messages, and acknowledging an incident stops the reminders without marking it resolved.

Yes. A maintenance window is an object with checks attached, so you put the whole release in one window. The checks keep running and the history stays honest; only the alerting stops.