Uptime check

What it asks you for

Checksrequired

How often to poll.

URLrequired

What to poll.

Healthy means

What a good response looks like.

What it tells the AI to do

Set up an uptime check.

- URL: {{ url }}
- Healthy: {{ healthy | default("HTTP 200") }}
- Checks: {{ schedule.human }} ({{ schedule.timezone }}), cron `{{ schedule.cron }}`

Create it with `automation create`, passing the cron and timezone verbatim.

Alert on transitions only: up to down, and down back to up. The recovery alert
matters as much as the failure one, because without it nobody knows whether to
keep looking.

Require two consecutive failures before declaring it down. A single failed poll
is usually the network between you and the target, and an alert per blip is an
alert nobody reads.

Include in the alert: the status or error, how long it has been failing, and
the time of the last good check.

• automation
• uptime
• monitoring
• ops

More automation skills

• Do this on a scheduleSet up a recurring automation that runs without you.
• Daily briefingA short scheduled digest of only what changed and matters.
• Watch and alert meCheck a condition on a schedule and speak up only when it fires.
• Daily task standupA morning digest of what is due, late, and waiting on you.
• Competitor digestA recurring roundup of what named competitors changed.
• Tidy DriveFile loose documents on a schedule, and report what moved.