VERIFIED PRODUCT EVIDENCE
These are founder-operated production investigations and controlled founder tests. They are not customer testimonials. Each story states the environment, the role of mttrly, the measured result, and what the evidence does not prove.
A controlled test shows that a specific workflow worked under recorded conditions. It is useful product evidence, but it is not the same as an unplanned customer incident, an independent testimonial, or a promise that every server will recover in the same time.
Who operated the server and whether the incident was intentional.
What happened from detection through approval and verification.
What cannot be concluded from the recorded result.
In a controlled founder test, mttrly detected an intentional Fail2ban stop, prepared a bounded remediation, and waited for separate human approval. The alert appeared after 6m14s; alert-to-resolution took 36s, including about 3s from approval to verified recovery, for an end-to-end cycle of about 6m50s.
Read the evidence and limitations →Create an account and install the outbound agent. Start with scoped server status, then use approval-gated actions when a change is needed. SSH remains available for bootstrap and emergencies.
Connect one server