Docs

From sign-up to your first alert in about two minutes.

Getting started

  1. Create an account and confirm your email. You get a 14-day Pro trial; no card needed.
  2. Enter a website. Observe checks its uptime every 5 minutes (set it anywhere from 1 to 60 under the monitor's Uptime settings) and watches its SSL certificate.
  3. Choose where alerts go. Email to you is on from the start; add Slack or anything else below whenever you like.
  4. A site is marked down after 2 failed checks in a row, so one network blip does not wake you. You get one alert when it goes down and one when it recovers.

Each monitor has more checks you can switch on: domain expiry, DNS records, DNS blocklists, broken links (crawl), sitemap and robots.txt, Lighthouse scores and heartbeats. Open a monitor and pick one from its sidebar.

Plans: Free covers 3 monitors, Pro 25 and Business 100. If you end up over your plan, the newest monitors pause until you upgrade or remove some; nothing is deleted. See pricing

Alerts and channels

Until you add a channel, alerts are emailed to the organization's owners and admins. Once you add any channel, alerts go to your channels instead, so add an Email channel too if you still want email. Manage them under Notifications & reports.

Email

Every owner and admin by default, or a fixed address.

Slack

An incoming webhook URL.

Discord

A channel webhook URL.

Microsoft Teams

An incoming webhook URL.

Google Chat

A space webhook URL.

Telegram

A bot token and chat ID.

PagerDuty

An Events API v2 routing key.

Opsgenie

An API integration key, US or EU. Opsgenie reaches end of support on April 5, 2027.

Jira Service Management

The API key of a JSM Operations API integration (Opsgenie's successor).

Pushover

Your user key and an application token.

ntfy

A server URL (ntfy.sh or your own) and a topic.

Webhook

Any HTTPS URL. Each request is signed so you can verify it came from Observe.

  • Per channel, choose which events it gets: downtime and recovery, SSL and domain expiry, DNS changes, Lighthouse regressions, crawl issues, missed heartbeats and more.
  • Send a channel alerts only during business hours, or only outside them, and limit it to monitors with certain tags or to one team.
  • Acknowledge an incident to stop further alerts until it resolves, or snooze a monitor for 1 hour to a week.
  • Opsgenie and Jira Service Management: setup, priority mapping, auto-close and migration notes are in the Opsgenie and JSM guide.
  • A channel that keeps failing for a day is paused and admins are emailed, so a dead webhook never hides an outage.

Status pages

Publish a public page showing the monitors you choose, under Org settings → Status page. It shows live up/down and SSL state, uptime history and recent incidents, with your title, logo, favicon, accent colour and language.

  • Visitors can subscribe by email to incidents and updates. Subscriptions are confirmed by email and can be cancelled from any message.
  • Post updates yourself, or let Observe post them automatically when a monitor goes down and recovers. Scheduled maintenance windows appear on the page too.
  • Make a page private: only people with its secret link can see it.
  • Every page also has a JSON feed (/s/your-page/json) and an RSS feed (/s/your-page/rss).

Cron heartbeats

Turn on Heartbeat for a monitor, set how often the job runs and a grace period, and call its URL when the job finishes. If a ping does not arrive in time, you are alerted.

# crontab: ping after the backup succeeds
0 3 * * * /usr/local/bin/backup.sh && curl -fsS -m 10 https://observe.live/heartbeat/YOUR_TOKEN

Append /starting before the job to measure how long it takes, or /failed to report a failure straight away.

Laravel apps can monitor every scheduled task at once with spatie/laravel-schedule-monitor. Point it at Observe with an API token that can write, then run schedule-monitor:sync after each deploy. Each task gets its own check and incident; the monitor's Heartbeat page shows the exact settings.

OH_DEAR_API_TOKEN=your-observe-token
OH_DEAR_API_URL=https://observe.live/api/v1/schedule-monitor/YOUR_MONITOR_ID/
OH_DEAR_MONITOR_ID=1
OH_DEAR_SITE_ID=1   # package v3

Servers

Add a server under Monitors → Add server with its hostname or IP. Turn on the checks you need:

  • Ping: round-trip time and packet loss every 5 minutes. Where ICMP is blocked, Observe falls back to a TCP connect on the ports you expect open (or 443, 80, 22).
  • Ports: list ports that must be open and ports that must stay closed; a change alerts you. An optional nightly full scan reports anything unexpected.
  • SSL: certificate validity and expiry on any port.
  • HTTP health check: a URL that must return the expected status code, and optionally contain some text.
  • Heartbeat: the server reports in on a schedule, so you hear about it when it goes quiet even if nothing can reach it from outside. No agent to install: a cron line is enough.
# crontab on the server: report in every 5 minutes
*/5 * * * * curl -fsS -m 10 -X POST https://observe.live/server-heartbeat/YOUR_TOKEN

Application health

Watch the inside of your app (database, queue, disk, cache) under Org settings → Health check. It reads the JSON format of Spatie Laravel Health, so a Laravel app needs no extra code.

  • Push: your app posts its results to your organization's URL on a schedule. If nothing arrives within the interval plus grace, you are alerted.
  • Pull: give Observe your health endpoint and it fetches it on the interval. An endpoint that stops answering, or answers with anything but valid health JSON, alerts you.
  • Each failing check alerts once, and you get one recovery notice when everything passes again.
# crontab: push your app's Spatie Laravel Health JSON every 5 minutes
*/5 * * * * curl -fsS -m 10 'https://your-app.example/health.json?fresh' \
  | curl -fsS -m 10 -X POST -H 'Content-Type: application/json' --data-binary @- https://observe.live/hc/YOUR_SLUG

Firewall allowlist

If your site sits behind a WAF or IP allowlist, permit our probe addresses so checks are not blocked as bots. The current list is published on the security page.

  • 104.28.161.245

Allow by IP address, not by User-Agent: our checks send an ordinary desktop browser User-Agent, so a WAF rule on it would also let real browsers through. In Cloudflare, add a WAF custom rule with the action Skip for these source IPs; elsewhere, add them to your firewall or IP allowlist.

Security: probe egress

Importing monitors

Moving from another tool? Go to Monitors → Add website monitor → Import, and bring your website monitors over with their names and check intervals:

  • UptimeRobot: paste your API key (a read-only key is enough).
  • Better Stack: paste an Uptime API token.
  • Oh Dear: paste an API token. HTTP monitors come with their uptime settings (keyword, expected status codes, method, timeout, headers), tags and upcoming maintenance windows.
  • Any tool: paste a CSV of url,label lines, or the CSV export from UptimeRobot or Better Stack as is.

You see the list first and pick what to bring over. Keys are used once and never stored. Monitors you already have are skipped, and your plan's limit is respected.

API and MCP

Everything above can be scripted. Create a token under Profile → API & data, then use the REST API or connect Claude, Cursor or any MCP client to ask about your monitors in plain language.

Read the API & MCP docs

Support

Email support@observe.live. We aim to reply within one business day, and sooner when alerts aren't arriving or a charge looks wrong.