Monitor every server, site and AI agent from one console.
SiteSonar is self-hosted monitoring for your whole stack. Push a metric from anything, set your tolerances once, and get alerted the moment a host, website, gateway or agent stops responding.
Everything in your stack, watched the same way
Push numbers from anything — a host, a database, an AI gateway, your own script. SiteSonar auto-discovers each metric and turns it into a live widget you can set tolerances on.
Servers & system
CPU, load, memory, swap and disk from a tiny self-updating agent. No agent on the box? A heartbeat URL works too.
Websites & uptime
Concurrent HTTP checks for up/down, response time and status codes — plus active TCP, UDP, ICMP, DNS and blacklist probes.
AI agents & gateways
Poll an API/AI gateway for up, latency and status, flatten its /metrics endpoint, or let an agent ping a dead-man's-switch heartbeat.
SSL certificates
Track days-to-expiry and chain/hostname validity for any domain, and get warned well before a certificate lapses.
Databases
MySQL throughput, connection pressure, buffer-pool efficiency and the slow queries that are actually hurting you.
Custom metrics
Push any numeric metric from any language or device. Each one auto-appears as a widget — zero server config required.
Get paged the moment an AI agent or gateway goes dark
Your AI workloads deserve the same vigilance as your servers. Watch gateways, workers and providers for outages, latency spikes and runaway cost — and know before your users do.
- Dead-man's switch. A worker or cron that stops checking in flips to offline automatically — no daemon required.
- Gateway health. Up/down, latency and status for any API or AI gateway, with its /metrics endpoint flattened into widgets.
- Token & cost tracking. Watch tokens/min, spend and error rate so a runaway agent is caught early.
- Confirm-after-N alerts. A single blip never pages you — breaches must persist across consecutive checks first.
Up and running in three steps
Connect a source
Drop in a one-line agent install, point the gateway poller at a URL, or just hit a /ping heartbeat on your own schedule.
Metrics auto-discover
Every numeric metric you send becomes a live gauge or sparkline automatically. No schema, no dashboards to build.
Set tolerances & relax
Choose warn/critical thresholds per metric and route alerts to email, SMS, Telegram, Pushover and more.
Start monitoring in minutes
Spin up SiteSonar on your own box and watch your whole stack — servers, sites, certs and AI agents — from a single console.