Downtime is a story.
Hear it from the start.
HeartGuard probes your endpoints every few minutes and tells you a plain story: status, response-time trend, incident timeline.
Add your first monitorFree: 5 monitors, 5-minute checks, email alerts.
api.acme.dev
99.98%
84ms p50
checkout.acme.dev
100%
132ms p50
docs.acme.dev
99.91%
610ms p50
Monitoring you will actually keep on.
Response-time trends, not just up/down
p50 and p95 per monitor catch degradation while the site still works - before customers notice.
Two failures before it screams
One bad check is a blip. Two consecutive failures open an incident and fire alerts; recovery closes it.
Clean-IP probes
Checks run from a single honest VM with SSRF guards and robots respect. Status and timing only - no content stored.
30-day uptime
99.98%
the screenshot number
incident timeline
- Jun 14 03:12 down - http 500
- Jun 14 03:26 recovered (14m)
- Jul 02 11:40 down - timeout
Degradation beats downtime
A slow endpoint is tomorrow's outage. HeartGuard tracks every check's TTFB so the trend line tells you first.
p95 612ms → alert at 800ms Simple enough to stay honest.
No distributed tracing. No multi-region probe matrix. One clean IP telling you the truth about status and speed.
POST /api/monitors { url, interval_s }
→ probe starts immediately -
Add your endpoint
URL plus optional expected status code.
POST /api/monitors {"url":"https://..."}
-
We probe on schedule
HEAD first, GET fallback; timing recorded per check.
every N minutes -> { status, ttfb_ms }
-
Two strikes open an incident
Blips stay quiet. Real outages page you and open the timeline.
fail + fail -> DOWN -> email/webhook
-
Recovery closes the story
First good check closes the incident with duration attached.
recovered after 14m
The anti-Datadog wedge.
| HeartGuard | UptimeRobot | BetterStack | |
|---|---|---|---|
| Free monitors | 5 | 50 (5-min floor) | suite-priced |
| Fast checks on free | 5-min | 5-min | - |
| Response-time tracking | yes | paid tiers | yes |
| No-false-alert logic | 2-strike incidents | single-check | configurable |
| Entry price | $0 | $0 crippled | $25/mo |
competitor positioning per public pricing pages, 2026
“It caught my API degrading over three days before it went down. UptimeRobot never showed me a trend line.”
Viktor Almeida
Solo SaaS founder
sample feedback
free tier
Five monitors. Zero excuses.
Your production endpoint deserves a pulse check every five minutes.
Add your first monitor