No ping?
Something's dead.

A dead man's switch is a monitoring pattern where your scheduled task pings a unique URL after every successful run. If the expected ping does not arrive within a configured time window, the monitoring service assumes the job has failed and sends an alert. The name comes from industrial safety devices that require an operator to actively hold a switch -- if the operator becomes incapacitated, the switch triggers a failsafe. In software, the same principle applies: a cron job, background worker, or scheduled task must actively report success. Silence is treated as failure. Unlike uptime monitors that check if a server is responding, dead man's switches detect jobs that silently stop running. Your server can have 100% uptime while a nightly backup or ETL pipeline has not executed in days.

DeadPing assigns each monitor a unique ping URL. After your cron job, background worker, or scheduled task completes, it sends an HTTP request (GET, POST, or HEAD) to that URL. DeadPing tracks the timing of each ping against the expected schedule, which can be a simple interval or a full cron expression. If a ping is overdue past the configured grace period, DeadPing fires an alert within 60 seconds via email, Slack, Discord, Microsoft Teams, PagerDuty, or OpenClaw. You can also send stdout, stderr, exit codes, and execution duration with each ping by using a POST request with the output as the body and query parameters for metadata. This gives you a complete audit trail of every job execution.

Uptime monitors like Pingdom or UptimeRobot actively check whether a server or URL is responding by sending periodic HTTP requests. They detect outages when something is broken and stops serving traffic. DeadPing solves the opposite problem: detecting jobs that silently stop running. Your server can show 100% uptime while a critical nightly backup, billing sync, or ETL pipeline has not executed in days. Nobody notices until data is stale, invoices are missed, or a customer reports the problem. Dead man's switch monitoring works by requiring each job to actively report success. If that report stops arriving, DeadPing knows something is wrong. Most production environments need both: uptime monitors for always-on services, and dead man's switches for scheduled jobs.

DeadPing offers three plans. Free: $0/month with 20 monitors, email alerts, 24-hour ping history, 1KB output capture, 1 public status page, and full API plus MCP access. No credit card required. Pro: $12/month with 100 monitors, all 6 alert channels (email, Slack, Discord, Microsoft Teams, PagerDuty, OpenClaw), 90-day history, 100KB output capture, full-text output search, progress tracking, bulk import from crontab and Kubernetes YAML, anomaly detection, and 5 public status pages. Business: $39/month with 200 monitors, 365-day history, 1MB output capture, alert grouping with incident management and auto-generated postmortems, and 10 status pages. All plans include the REST API and MCP server for AI assistant integration.

DeadPing supports six alert channels. Email is included on all plans including Free. Slack, Discord, Microsoft Teams, PagerDuty, and OpenClaw are available on Pro and Business plans. All channels fire within 60 seconds of a missed ping, so you find out about failures almost immediately. Each monitor can be configured to alert on specific channels, so you can route critical job failures to PagerDuty while sending less urgent notifications to Slack or OpenClaw. The Pro plan includes anomaly detection alerts that notify you when a job's duration, frequency, or error rate deviates from its learned baseline. The Business plan adds alert grouping that consolidates related failures into a single incident notification.

Yes. Send a POST request to your monitor's ping URL with stdout and stderr as the request body. Pass the exit code and execution duration as query parameters. DeadPing stores the output alongside the metadata and displays it in your dashboard for each ping. Storage limits scale with your plan: 1KB on Free, 100KB on Pro, and 1MB on Business. Output is truncated at the byte level if it exceeds the limit, not silently dropped. Pro and Business plans include full-text search across all captured output, so you can find a specific error message across thousands of pings. You can also track multi-step job progress by passing progress, total, and message query parameters with each ping.