OpenClaw Heartbeat Monitoring: How to Set Up 30-Minute Health Checks

Your agent runs while you sleep. Without monitoring, it's completely silent about problems. A payment failed three days ago? You'll find out when the service cuts off. A domain renewal is due? You'll find out when the site goes down. An important email arrived at 2am? It waits until you happen to check.

OpenClaw heartbeat monitoring solves this. Every 30 minutes, your agent does a quiet health check — emails, calendar, services, security posture — and only surfaces what you actually need to know. Most checks find nothing. When something matters, it tells you immediately.

---

What Heartbeat Monitoring Is

A heartbeat in OpenClaw is a repeating automated check that runs at a set interval. Unlike a one-off cron job, it's designed to run continuously and indefinitely.

The name comes from the concept of a pulse check — something that confirms the system is alive and healthy. In practice, it's more useful than that: it's an always-on background scanner that looks for things you care about.

The default interval used by experienced OpenClaw operators is 30 minutes — frequent enough to catch time-sensitive issues, not so frequent it generates noise.

---

What to Monitor in a Heartbeat Check

The most effective heartbeat configurations monitor four categories:

1. Email and Communications

• New emails from priority senders (clients, family, important services)
• Billing alerts and payment failures
• Time-sensitive requests that need a response within a few hours

2. Calendar and Schedule

• Meetings starting within the next hour
• Scheduling conflicts
• Events added by others to shared calendars

3. Services and Infrastructure

• Domain expiration warnings
• SSL certificate expiration
• Uptime monitoring for services you run
• Subscription renewals and billing cycles

4. Security Posture (Optional)

• Unexpected changes to config files
• Failed authentication attempts
• New SSH connections to your server

---

Setting Up a 30-Minute Heartbeat

Send this prompt to your OpenClaw bot to establish the monitoring loop:

Set up a heartbeat that runs every 30 minutes. Each run should:
1. Check my email inbox for anything from my priority contacts or billing-related messages
2. Check my calendar for upcoming events in the next 2 hours
3. Check for any domain renewals due in the next 30 days

If nothing urgent is found, stay silent. If something needs my attention, send me a brief message with exactly what you found and what (if anything) you recommend I do.

The "stay silent unless something matters" instruction is important. A heartbeat that messages you every 30 minutes with "all clear" becomes noise you start ignoring. A heartbeat that only messages you when something needs your attention stays useful indefinitely.

---

Real Examples of What Heartbeat Catches

Two documented examples from a user running heartbeat monitoring for 50+ days:

Netflix payment failure: The agent caught a payment failure during a routine email scan. It sent a brief alert: "Netflix payment failed — billing email received at 11:42pm. Action required." The user paid within five minutes. No service interruption.

Without heartbeat monitoring: this sits in the inbox until the user happens to check email, potentially after Netflix has already suspended the account.

Domain renewal: The agent found a domain renewal notice in the inbox and included it in a heartbeat report. The user had 12 days to renew before expiration. They renewed the next day.

Without heartbeat monitoring: domain renewals frequently slip, especially if you own multiple domains or the renewal notice gets buried under other email.

The pattern here: heartbeat monitoring doesn't do anything you couldn't do manually. It does it reliably and without forgetting.

---

Configuring the Heartbeat Scope

The heartbeat scope should match what you actually care about. Don't monitor everything — monitor what matters.

Light version (good starting point):

Every 30 minutes, check my email for anything urgent or from my priority contacts list. Only alert me if there's something that needs action today.

Standard version:

Every 30 minutes:
- Email: flag anything urgent, billing alerts, or from [list key contacts]
- Calendar: warn me 90 minutes before any meeting
- Services: check if [specific domain] is still resolving correctly

Silent unless you find something. If you find something, one message with the relevant details.

Full version (for power users):

Every 30 minutes:
- Email: priority contacts, billing, time-sensitive requests
- Calendar: upcoming events, conflicts, new additions
- Infrastructure: domain expiry, SSL expiry, service uptime
- Security: failed auth in last 30 min, config changes, new SSH sessions
- Finance: weekly API spend summary (Sundays only)

Stay silent unless something is flagged. If flagged, immediate alert with action recommendation.

---

Combining Heartbeat With the Daily Security Audit

Heartbeat monitoring and the daily security audit serve different purposes:

Run both. The heartbeat keeps you informed about what's happening in your world. The daily audit keeps your security baseline intact.

# Two separate cron configurations
# Heartbeat: every 30 minutes, check email/calendar/services
# Daily audit: every morning at 9am, security posture check

---

Sample Heartbeat Configuration Prompt

Here's a complete, copy-paste ready prompt for setting up heartbeat monitoring:

Set up ongoing heartbeat monitoring that runs every 30 minutes, continuously. 

Each heartbeat should silently check:
1. Email inbox — flag anything from VIP contacts, any billing failures, and anything marked urgent
2. Calendar — warn me 60 minutes before any meeting I haven't acknowledged
3. Domain renewals — alert if any domain in my config is expiring within 30 days

Rules for alerting:
- If nothing needs attention: stay completely silent
- If something needs attention: send one concise message listing exactly what was found
- Format: short bullet list, what it is, when it needs action
- Do NOT include irrelevant context or lengthy explanations

Start the heartbeat immediately and confirm it's running.

Adjust the specific checks to match your integrations and what you actually care about monitoring.

---

Why "Silent Unless Urgent" Matters

The most important design principle in heartbeat monitoring is aggressive filtering. An agent that messages you frequently with updates you don't need is an agent you start ignoring — which means you'll miss the important alerts too.

The goal is a heartbeat that runs dozens of times per day and produces output maybe twice per week, but when it does produce output, you know it's worth reading.

Experienced OpenClaw users describe this as the agent "stopping being a chatbot and becoming a system." You stop thinking about it and it just catches things that would otherwise slip through.

[→ See also: How to Set Up a Daily Security Audit Cron Job in OpenClaw] [→ See also: OpenClaw Email Security: Why Draft-Only Mode is Non-Negotiable]

---

Key Takeaways

• Heartbeat monitoring runs a silent check every 30 minutes while you do other things. It only messages you when something needs attention.
• Monitor four categories: email/communications, calendar/schedule, services/infrastructure, and (optionally) security posture.
• The "silent unless urgent" rule keeps the heartbeat useful long-term. Constant all-clear messages become noise.
• Real-world catches include payment failures, domain renewals, and calendar conflicts — things that would otherwise slip through the cracks.
• Run heartbeat (30-min) alongside a daily audit (9am) for both real-time awareness and security compliance checking.
• Keep the scope focused on what you actually care about. More monitoring isn't always better — relevant monitoring is.