Your API is down. You found out from Twitter.

Kernelworks watches your endpoints and screams in your ops channel the second one breaks. Open source. Zero dependencies.

Star on GitHub

What it does

Screams first

Slack, Discord or PagerDuty alert within one second of a bad status code. Before the tweets.

--alert slack

Zero deps

One binary. No node_modules black hole.

4.2 MB

Watches latency

Not just up or down. p95 creeping past your budget pings you too.

--p95 300ms

Runs anywhere

Laptop, CI, or a $4 VPS. It is one process that does one thing.

--daemon

Config in one file

A tiny YAML. No dashboard to click through, no account to make.

kernel.yaml

Replays failures

Keeps the last failed request and its full response so you debug the real thing.

--keep 50

Three steps

1
Install it.
npm install -g kernelworks
2
Point it at your API.
kernelworks watch api.example.com
3
Wire the alert.
kernelworks alert slack #ops

Developers say

"Caught a bad deploy before our status page did. Genuinely."

@devops_rina

"Replaced a $200/mo tool with a 4MB binary. My CFO loves me."

@budi_builds

"The one alert I do not mute. It only fires when it matters."

@sre_katherine

Stop finding out last.

npm install kernelworks
Copied!