Incident bot for Slack
Probes the sites every 10 minutes and posts each incident to Slack with what it saw, who owns it and buttons to act.
The problem
A test site ran an old backend for hours. Every dashboard looked fine, and nobody was told.
What I built
- A probe every 10 minutes, and a heartbeat on separate runners that alarms when the probe itself stops.
- Checks for certificates, missing API gateway routes, slow pages and the login journey.
- Incident cards with Open, Mute, Hand over and Resolve buttons, and escalation when nobody takes one.
- A drill target to prove it works end to end.
In numbers
- 10 min
between probes
- 2
heartbeats an hour
- 9 min
first drill, down to cleared
Stack
GitLab CI schedulesBashSlack appNext.js relay
Want something like this? Tell me about your project or hire me on Upwork.