Troubleshooting
The usual causes, in the order to check them.
The server shows as offline
- sudo systemctl status secai-agent, then sudo journalctl -u secai-agent -n 100.
- Egress: the agent needs HTTPS to secai.techsteps.ae. Test with curl -sI https://secai.techsteps.ae/health from the server.
- DNS: a resolver failure looks like an outage. The agent keeps the last known address; check /etc/resolv.conf if it never resolved.
- Time: a clock far off UTC can make signed requests fail. Check timedatectl.
I locked myself out
Use the "Let me in" link in the alert email, or the Unblock action on the server page. Unblocking never needs approval and works while automation is paused. Your own address should be marked trusted so it cannot be blocked again.
A fix broke something
The agent rolled it back on the server from the copy it kept, and emailed what was attempted. That kind of change on that server now waits for approval; re-enable it from the server page when you are ready. Undo any applied change from its record.
An update did not take
The previous binary is at /usr/local/bin/security-agent.prev and update attempts are logged in /var/log/securityemployee-agent-update.log. The next check is within a day; sudo systemctl restart secai-agent triggers one at start.
Something on this page is wrong
Tell us through the Contact page, secai.techsteps.ae/contact. A sentence here that the code no longer makes true is a defect.
Reviewed against the implementation on 2026-09-20. Something wrong? Tell us.